The Invention of Printing in China and Its Spread Westward

278 60 16MB

English Pages 293 [343] Year 1955

Report DMCA / Copyright


Table of contents :
Introduction to Second Edition
Thomas Francis Carter
1 The Invention of Paper
2 The Use of Seals
3 Rubbings from Stone Inscriptions
4 Religion and the Demand for Printing
5 The Significance of Block Printing, Ink, and the Method Used
6 The Beginnings of Block Printing
7 The Empress Shotoku of Japan and Her Million Printed Charms
8 The Diamond Sutra of 868, the Oldest Extant Printed Book
9 The Printing of the Confucian Classics Under Fêng Tao, 932–953
10 The High Tide of Chinese Block Printing, 960–1368
11 The Printing of Paper Money
12 Commerce in Thought and Wares Along the Great Silk Ways
13 Paper’s Thousand-Year Journey from China to Europe
14 The Printing of the Uigur Turks
15 Islam as a Barrier to Printing
16 Contacts Between China and Europe in the Mongol Empire
17 Persia, Crossroads Between East and West
18 Block Printing in Egypt in the Period of the Crusades
19 Playing Cards in the Westward Movement of Printing
20 The Printinc of Textiles
21 Block Printing in Europe
22 The Invention of Movable Type in China
23 The Wide Use of Movable Type in Korea
24 The Pedigree of Gutenberg’s Invention
Chart: Paper and Printing, Evolution in China and Spread Westward
List of Words
Recommend Papers

The Invention of Printing in China and Its Spread Westward

  • 0 0 0
  • Like this paper and download? You can publish your own PDF file online for free in a few minutes! Sign Up
File loading please wait...
Citation preview

Thomas Francis Carter




Revised by





Copyright. 1955, by The Ronald Press Company

Copyright, 1925, by The Ronald Press Company

All Rights Reserved The text of this publication or any part thereof may not be reproduced in any manner whatsoever without permission in writing from the publisher.

Library of Congress Catalog Card Number: 55-5418 PRINTED







To PAUL PELLIOT Membre de l’lnstitut, Professor of the Languages, History, and Archaeology of Central Asia in the College de France, the master mind of Chinese historical research; whose example, whose writings, and whose revision of the manuscript have made possible such measure of accuracy as this work can claim

Digitized by the Internet Archive in 2019 with funding from Kahle/Austin Foundation





Introduction to Second Edition Thomas Francis Carter



The Invention of Paper.


The Use of Seals


Rubbings from Stone Inscriptions


Religion and the Demand for Printing

........ ..... ....


The Significance of Block Printing, Ink, and the Method Used.



The Beginnings of Block Printing ..



The Empress Shotoku of Japan and Her Million Printed Charms.


The Diamond

Sutra of 868, the Oldest Extant Printed





The Printing of the Confucian Classics Under Ff:ng Tao,





The High Tide of Chinese Block Printing, 960-1368




The Printing of Paper Money.








Commerce in Thought and Wares Along the Great Silk •



Paper’s Thousand-Year Journey from China to Europe




The Printing of the Uigur Turks.140


Islam as a Barrier to Printing.150


Contacts Between China and Europe in the Mongol Empire


Persia, Crossroads Between East and West.168


Block Printing in Egypt in the Period of the Crusades




Playing Cards in the Westward Movement of Printing




The Printinc of Textiles.



Block Printing in Europe









The Invention of Movable Type in China







The Wide Use of Movable Type in Korea.223


The Pedigree of Gutenberg’s Invention



Chart: Paper and Printing, Evolution in China and Spread Westward. .










List of Words








Thomas Francis Carter.

Frond (piece

Wooden Stationery of the Han Period.


Early Paper.


Chinese Seals and Seal Impressions



Primitive Tibetan Charm Prints.


Clay Containers for Tibetan Charm Prints



Rubbing Being Taken from a Stone Inscription



Metal Stamp for Making Figures of Buddha


Fragment of a Roll of Thin Paper with Stamped Buddhas


Paper Stencil or Pounce.


The World’s Oldest Printing.


The Diamond Sutra of 868, the World's Oldest Extant Prin ed Hook


Block Print Presented in Payment of a Vow


Tenth-Century Woodcut from Tun-huang





Evolution of the Chinese Book.


Pages from a Printed Book of the Sung Dynasty .


One of the Earliest Printed Books of Japan


Printed Sutra of 956



Buddhist Woodcut of the Sung Dynasty Bronze Block for Printing Currency in the Time of Kublai Khan 1287).


Buddhist Sutra.


Page from the Sanskrit Diamond Sutra


Leaf from a Sanskrit Book Printed to Imitate the Indian Pothi


Bit of Tangut Printing.



Fragment of a Printed Sutra in the Mongol Language in Square (’Phags-pa) Script. Letters from the Mongol Rulers of Persia to Philip the Pair of Prance vii

121 168



Red Seal Impression from the Letter of 1289 to Philip of France





Oldest of the Egyptian Block Prints.169 Old Chinese Playing Card.184 Revolving Wheel Typecase. Wooden Types and Impressions of the Early Fourteenth Century


184 .



Early Korean Metal Types.185


inventions that spread through Europe at the beginning

of the Renaissance had a large share in creating the modern world. Paper and printing paved the way for the religious reformation and made possible popular education. Gunpowder leveled the feudal system and created citizen armies. The compass discovered America and made the world instead of Europe the theater of history. In these inventions and others as well, China claims to have had a conspicuous part. The purpose of the present work is to investigate the truth of this claim in the one domain of printing. The restlessness of the tribes of Central Asia during the early centuries of our era brought several hundred years of anarchy in China, corresponding to the Dark Ages in Europe; but as these barbarian migrations did not cause quite such a complete rooting up of classical civilization in Eastern Asia as they did in the West, China quickly recovered and was earlier ready for those inventions which came into Christendom with the beginning of the Renaissance. Marco Polo’s record shows us a China whose civilization already in the thirteenth century had come to full bloom and had advanced very much further than that of contemporary Europe. When Europe was ready for the new life, she found in the Arabic empire and at Constantinople reservoirs ready at hand where the lore of her own classical world had been stored away, and to these reservoirs she turned with a real thirst. But with the classic lore there was a certain new element that also entered Europe from the East—an essentially modern spirit of invention and practical dis¬ covery. The mediators of the inventions that reached Europe at this time were the Arabs and the Empire of the Mongols. But the in¬ ventors were neither Arab nor Mongol. There seems to be good reason to believe that certain processes that had been gradually evolved in China, when joined with the recovered civilization of Greece and Rome, had much to do with starting Europe forward on IX



her course of progress, a course to which the classics alone could never have led. It is the glory of European genius, newly awakened, that it was able to seize these discoveries, dimly seen in Eastern Asia and in some cases but dimly understood in the land of their birth, and to make of them the basis for a civilization of which their dis¬ coverers could never have dreamed. Pre-eminent among these inventions of China, on account of their influence both in Eastern Asia and in Europe, stand paper and printing. The invention of paper has already received considerable attention. The scientific study of the subject in the West was begun by Dr. Friedrich Hirth,1 who held for many years the chair of Chinese at Columbia University, and its popular presentation has been carried forward in general histories.

The facts concerning

China’s part in the invention of printing, on the other hand, have been almost unknown to European scholarship, except in a few of their larger outlines. No historical research, however, can lay claim to complete orig¬ inality, and this study of Chinese printing may be considered a com¬ pendium of the researches of a multitude of scholars—Chinese, Japa¬ nese, and Western scholars of many centuries—correlated with cer¬ tain of the results of excavations in Turkestan and in Egypt. The bibliography indicates the main sources, and indicates also the debt of gratitude felt by the author to all these investigators, the results of whose labors have been freely borrowed. On the other hand, the gathering together and correlating of this source material from dif¬ ferent ages and different parts of the world has been largely a virgin field. It is this which has made the work at the same time difficult and inspiring. Apparently, the first mention in European literature of the Chinese invention of printing dates from the year 1546, when the Italian historian Jovius, from an examination of certain printed books brought from Canton by Portuguese travelers and presented by the King of Portugal to the Pope, came to the conclusion that European printing was derived from China.2 In the eighteenth cen1. Actually, according to Laufer, 1927:74, Hirth’s article was based on earlier papers by Edkins, 1867:67-68; and Wylie, 1867 and i922:xiv—xv. 2. For full quotation from Jovius, see Chapter 16, note 4.



tury Phil. Couplet in the British Encyclopedia, writing evidently on the authority of Roman Catholic missionaries, assigned the year 930 as the date of the Chinese invention.

Gerard Meerman in his

Origines Typographicae in 1765 told of early Chinese printing, basing his statement on Arabic authority.3 A further study of the subject from Chinese sources was made by Jules Klaproth4 in 1834 and by Stanislas Julien in 1847. The results of Julien’s work were published in a short article in the Journal Asiatique, which, in spite of inaccuracies, has formed the basis of practically all that has been written on the subject in the West up to the present. A letter from Thomas T. Meadows to Lord Elgin, pub¬ lished as part of a paper by Lord Robert Curzon in the Miscellanies of the Philobiblon Society of London in 1858, contains what is probably the best account of the Chinese invention of block printing that has appeared in any European language even down to our own day, but unfortunately this letter has been hidden away in a littleknown publication and in an article the balance of which is of doubt¬ ful value, and it has apparently escaped the attention of later writers.6 After 1858 little if any independent work devoted to printing in China appeared in any European language until 1923, when Dr. Her¬ mann Hiille of Berlin published in a fifteen-page booklet a clear summary of the history of Chinese typography and its development in Korea, based partly on Julien and Sir Ernest Satow and partly on independent research in Chinese sources.6

The writer had the

privilege of working for some months under the expert direction of Dr. Hiille, who is in charge of the Chinese department in the State Library at Berlin, and who very kindly placed all his source material at the writer’s disposal. 3. Meerman’s authority is “The Histona Sinensis of Abdalla, written in Persic in 1317, which speaks of it [printing] as an art in very common use. The reference is probably to Banakatl, whose work is quoted from that of Rashid-eddin. See Chapter 17, note 19. For this and other early European statements about Chinese printing, see Thomas, 1810:72. 4. For this and subsequent references see the Bibliography. 5. A very brief but accurate notice of the invention of printing in China is contained in a book entided China by J. F. Davis, 1857-II, i73~74- ^ believe the same passage occurs in an earlier edition of the work published in 1836. 6. E. H. Parker, 1894-95:119-20, however, wrote some useful notes on the origin of printing.



Meanwhile an article by Sir Ernest Satow on the history of early printing in Korea and Japan was published in 1882 in the Transac¬ tions of the Asiatic Society of Japan, and has remained to the present the main source of what is known in the West on that subject. Modern scholarship in Japan and China has produced at least three works7 which gather up the main historical facts concerning the history of printing in their respective countries, the Japanese work, as is natural, dealing somewhat with Chinese sources and more fully with those of Korea, as well as with the Japanese de¬ velopment of the invention. Unfortunately, these books are avail¬ able only in Japanese and Chinese. All are brief but are far more complete than the short sketches mentioned above that have ap¬ peared in European languages. These articles and books in five of the world’s leading languages have been used freely in the preparation of the present work, both for the actual information contained and more especially for their references to earlier Chinese literature. Another important source has been the great Chinese encyclo¬ pedias, especially the T’u shu chi ch’eng, published in 1726-28, and the Ko chih ching yuan, published in 1735. These too have been valuable, largely on account of their quotations from earlier works.8 Unfortunately, while new improvements in the art of writing, such 7. Asakura, Yeh Te-hui, and Liu-an; see the Bibliography. 8. Not every quotation from Chinese encyclopedia or other source book can be carried back to its ultimate source, as would be preferable in the interest of scientific accuracy. In key passages, however, on which the framework of the history depends, every effort has been made thus to get back to the original statement and to compare variant editions. In other cases, where use has been made of secondary sources, and in the very few cases where the translations of other European scholars have been accepted, the secondary source as well as the original has been noted. It should be observed that the Chinese encyclo¬ pedias and other source books used consist almost wholly of verbatim quota¬ tions from earlier works rather than paraphrases or summaries, and that, while the possibility of copyists’ errors is not thus altogether excluded, this method greatly diminishes the likelihood of such error. Laufer, 1927:71 does not agree as to the reliability of Chinese encyclopedias. But there are instances when the encyclopedias, quoting from earlier editions, are closer to the originals than the modern editions of well-known books. See Goodrich and Ch’ii T’ung-tsu, 1949:149; and Balazs, 1953:206-8.



for instance as the invention of the hair pen and the invention of paper, have called forth a voluminous literature of antiquarian re¬ search by Chinese writers, printing has as a rule been taken for granted and sparsely mentioned. Calligraphy has been considered the work of artists, printing that of artisans. However, by supple¬ menting such direct references as have been found with many in¬ direct references, it is possible to gain a fairly clear picture of the early history of the art, at least as clear a picture as we have of early European block printing, which grew up equally in the dark. A further source, and that which gives us our most certain in¬ formation, is archaeology. The desert air of Chinese Turkestan, like that of Egypt, has preserved intact the memorials of ancient civiliza¬ tion, and the researches of British, French, German, Russian, and Japanese expeditions have made it possible to reconstruct the history and daily life of these western outposts of China during the first thou¬ sand years or more of our era. One result of this research has been clear testimony to the accuracy of the Chinese records of the period. Another result, bearing more directly on the subject in hand, has been the discovery in different parts of Turkestan and its border¬ lands of a large number of block prints and block books of varying date which shed light both on the progress of the art of printing in China and on its westward course. Excavations in Egypt also have revealed the products of a hitherto unsuspected block printing ac¬ tivity continuing through the time of the Crusades, the significance of which must still be regarded as something of a mystery, but which may eventually lead the way toward the discovery of the connection between the block printing of the Far East and that of Europe. As indicated above, it is not only to books that the writer is in¬ debted. A far more personal debt must here be acknowledged. The keenest pleasure in the preparation of the work has been the counsel, guidance, and criticism—and the friendship—of some of the world’s leading scholars in the realms of Chinese, Central Asiatic, and Arabic history. In this work nationality has been forgotten. In Berlin and Vienna, as well as in Nanking, Paris, and London, unfailing kind¬ ness and cooperation have been met.



The expert guidance of Dr. Albert von Le Coq, given freely day after day in the study of the Turfan discoveries at Berlin, the inspira¬ tion given by Dr. Adolf Grohmann of Prague in the study of the block prints of Egypt at Vienna, the help afforded by Mr. Arthur Waley and Mr. Lionel Giles in the examination of the Tun-huang finds at the British Museum, the well-nigh perfect library assistance afforded by Dr. Hermann Hiille of Berlin, and the patience of my colleagues at Columbia University, Professor Lucius C. Porter of the Chinese Department, Professor A. V. Williams Jackson of the IndoIranian Department, Professor Richard J. H. Gottheil of the Semitic Department, and Professor William L. Westermann of the Depart¬ ment of Ancient History, in reading the manuscript and making valuable suggestions, all place the writer under a debt of gratitude such as can never be repaid. But deepest of all is my obligation to Professor Paul Pelliot of the College de France. Professor Pelliot has set a new standard of ac¬ curacy and acumen in Chinese research to which all investigators are indebted.

His researches in literature and in archaeology have

furnished a mass of facts on which many of the conclusions of this book are based. The debt of the writer to Professor Pelliot goes further, for Professor Pelliot has patiently gone over the first draft of the manuscript chapter by chapter, has gradually introduced the writer to more clear-cut and accurate methods of Chinese research, has made on almost every page suggestions and corrections which the writer has sought to follow up and incorporate, and has given freely of his store of historical understanding. In such a work as this, it is impossible to acknowledge one’s debt to all who have freely rendered assistance, but to the following, who, in addition to those already mentioned, have given largely of their time and their expert knowledge, a word of gratitude must be ex¬ pressed: Dr. Vasseely Alexeiev, professor of Chinese Philology, University of Leningrad (Petrograd) ; Mr. Laurence Binyon, curator of Oriental Art, British Museum; Professor Edward G. Browne, Department of Arabic, Pembroke College, University of Cambridge; M. Henri Cordier, Membre de l’lnstitut, professor of Chinese His¬ tory in the Ecole des Langues Vivantes, Paris; Pere Henri Dore, author of Superstitions en Chine; Dr. Erich Hanisch, professor of



Chinese, University of Berlin; Dr. Sven Hedin, head of the Swedish expeditions of exploration in Central Asia; Mr. John Hefter, librarian of Chinese books, Columbia University Library, New York; Dr. Friedrich Hirth, former Dean Lung professor of Chinese, Columbia University; Mr. T. S. Hsu, of the faculty of Chinese History, Peking University; Mr. Homer B. Hulbert, author of The History of Korea; Mr. Y. F. Hung, head librarian, National Southeastern University, Nanking; Rev. William C. Kerr, American Presbyterian Mission, Seoul, Korea; Dr. Sten Konow, professor of Sanskrit, University of Kristiania; Dr. Berthold Laufer, curator of anthropology, Field Museum of Natural History, Chicago; Mr. S. Y. Li, acting librarian of the Chinese collection, Library of Congress, Washington; Dr. David S. Margoliouth, professor of Arabic, Oxford University; Dr. Bernhard Moritz, professor of Arabic, Seminar fur Orientalische Sprachen, Berlin; Dr. F. W. K. Muller, director of the Chinese and Indian departments, Museum fur Volkerkunde, Berlin; Professor Rudolf M. Riefstahl, Department of Fine Arts, New York Univer¬ sity; Dr. Clementz Scharschmidt, professor of Japanese, Seminar fur Orientalische Sprachen, Berlin; Dr. Theodor Seif, curator of Arabic papyri and papers in the Erzherzog Rainer Collection, Austrian National Library, Vienna; Dr. Adolf Stix, curator of European in¬ cunabula, Austrian National Library, Vienna; Dr. Walter T. Swingle, chairman of Library Committee, United States Department of Agriculture, Washington; Dr. Clark Wissler, curator of an¬ thropology, American Museum of Natural History, New York. Grateful acknowledgment should also be made of the sources from which illustrations have been received. These are in the main from original photographs taken in the museums where the objects are preserved. The writer is specially indebted to Mr. Waley and to Dr. von Le Coq for their courtesy and assistance in obtaining photographs from the British Museum and the Museum fur Volkerkunde. Where illustrations are reproduced from other books, ac¬ knowledgment is made in abbreviated form beneath the illustration concerned. The romanization of Chinese words is that of Giles, which, in spite of serious drawbacks, is the one most generally used among scholars. Exceptions are made of the names of provinces and large



cities like Peking, where the post office romanization has been fol¬ lowed. The names of those dynasties that are easily confused in Giles’ romanization are here spelled in the more easily recognized form, Ch’in, Tsin, and Kin. The hope with which this book goes forth cannot be better expressed than in the words of the Chinese writer Tai Tung, who wrote and had printed during the thirteenth century a book on the history of Chinese writing: Were I to await perfection, my book would never be finished, so I have made shift to collect the fruits of my labors as I find them. It was said by the Master, “In preparing the governmental notifications, P’i Shen first made the rough draft; Shih Shu examined and discussed its contents; Tzu-yii, the manager of foreign intercourse, then made addi¬ tions and subtractions; and finally Tzu-ch’an of Tung-li gave them the proper elegance and finish.” Such a rough draft is the present work. For the examination and discussion of whatever truth it con¬ tains, it awaits the judgment of a master-mind, . . . one whose wise and lofty spirit will lead him, without looking down upon the author, to . . . correct and suppress where the text is in error, to add where it is defective, and to supply new facts where it is altogether silent.9

T. F. C. 9. Tai T’ung, Liu shu \u, concluding paragraph. Translation adapted from L. C. Hopkins, 1881:60-61. The quotation from Confucius is from the Analects, book 14, chapter 9.




ago Dagny Carter inquired if I would undertake a

revision of this book. I hesitated, as I did not wish to tamper with the fresh, almost eager style of Thomas Francis Carter’s writing by introducing changes and additions resulting from the new knowl¬ edge gained since his untimely death in 1925. But these hard facts were compelling: the book was out of print, and it clamored for cer¬ tain revisions if brought out again. Working on the book has been almost as exhilarating for me as it seems originally to have been for Dr. Carter. It has meant check¬ ing on a thousand points, and has sent me scurrying about our library to consult works I have either rarely or never consulted be¬ fore, communicating with authorities in near or far-off places, and visiting certain collections of rareties which previously I had not seen: a certain private collection in New York, the treasure room of Horyu-ji in Nara, the Tun-huang treasures at the Museum of Central Asian Antiquities (New Delhi), the Bibliotheque Nationale, and the British Museum, and the Gutenberg Museum at Mainz, for example. We have advantages in the United States at present which did not exist in Dr. Carter’s day. He was forced by the smallness of our Chinese library holdings to lean heavily on a few encyclopaedic works. We can now go back of them in almost every case to the originals from which their compilers quoted.

Further, his own

volume played its part in stimulating new researches by Chinese, Japanese, European, and American scholars.

Add to these ad¬

vantages recent discoveries by archaeologists. For example, in 1931 Folke Bergman found in central Asia fragments of paper in associa¬ tion with wooden manuscripts, the dates of which he was unable to ascertain at the site. Seventeen years later Mr. Lao Kan determined them to be equivalent to the years between


93 and 98. The

paper is not surely of the same half decade, but it is certainly older than anything known a generation ago. In regard to early printing, xvii



savants in various lands have turned up four references, unknown to Carter, in the Chinese and Japanese literature of


835 to 907; and

these show that secular and Taoist circles as well as Buddhist were active in utilizing the new technique. Early printed pieces (one dated 956, the other 975) have also come to light near the east coast of China, indicating that the lower Yangtze valley as well as Szechuan and the capital area in Honan were centers of the new art. Carter’s book this was, and Carter’s it should remain. He laid it out on grand lines and as such it was a contribution to literature and to knowledge. My sole desire has been to correct a few mistakes and bring it up to date. I owe many acknowledgments for assistance rendered. They are given, I trust adequately, in the bibliography and in the notes. L. C. G. Columbia University, New York March, 1955


spring of 1921 North China was threatened with another famine, and the call had gone out for volunteers to help in or¬ ganizing emergency work and relief. As several times before, Tom Carter had responded, and it would probably not have been recorded here but for a small event that happened at this time. On his way to Shantung he found time to read on the train W. J. Clennell’s book The Historical Development of Religion in China, and in his reading he came across the statement that inspired the opening paragraph in his Introduction to the present volume. He asked himself, Has the Chinese origin of these four epoch-making inventions been proved? How much is actually known about the beginnings of paper, printing, gunpowder, and the compass ? After the famine was over, he returned to the schools that had been his main work for the past eight years. But the spark from the passage in Clennell’s book remained in his mind until the question took possession of all his waking hours. The scholar he was intended to be had been born. In the summer of 1922, after spending another spring in bringing famine relief by superintending the making of three hundred miles of dirt roads with famine labor and dispensing half a million dollars in relief to thousands of destitute homes, he left his chosen work to begin his search for facts about the inventions mentioned in Clennell’s book. Thomas Carter came to China for the first time in 1906. He was then two years out of Princeton and was making a world tour with his Princeton contemporary Norman Thomas and two other friends. That this was not to be a usual sightseeing tour soon became ap¬ parent. The world problems and many new interests that met him at every turn, as well as the ever-present thought of choosing a life In the




work, made it increasingly difficult for him to follow the scheduled itinerary. In Nanking he left his travel companions to go into the interior to visit the homes of two cousins who had come to China as missionaries. As there were no train connections between Nanking and Huaiyuan in those days, he joined a party of carpenters who were making the one hundred and fifty mile trek on foot or on donkey-back. Friends in Nanking provided him with food and bedding and taught him a few Chinese words, which he had to make the most of with his fellow travelers who knew not a word of any Western language. This gave him his start. Although he did not plan to stay in Huaiyuan more than a few weeks, as soon as he arrived he secured a teacher and began his study of the Chinese language. He made ex¬ ceptional progress the three months he was there, and after return¬ ing to America for his graduate studies, he kept up a correspondence in Chinese characters with his teacher in Huaiyuan. When he re¬ turned to China in 1910 with his bride to take a permanent position, he was able immediately to use the language he had begun studying in 1906. It was the following ten years spent in a small Chinese town developing and superintending a circuit of city and country schools in close association with the Chinese that gave to him his rare under¬ standing and appreciation not only of the Chinese themselves but also of that background which has made them what they are, the world’s most patient, tolerant, and truly cultivated people. He used to say, I came to teach but I stayed to learn.” The eagerness of the young for new experiences remained one of his salient characteristics. He had no sooner arrived in China in 1910 than he took up as his most absorbing avocation a study of Chinese history, both from books and from conversations with the scholars he met who love nothing better than a discourse on their ancient lore. In the Chinese language he continued to prove himself a scholar of unusual merit. To this study, with its great difficulties, he brought a joyousness of spirit that made play of what might otherwise have been a tiresome, nerve-racking task. When Tom Carter and 1 left China in the summer of 1922 on a



leisurely trip to Europe, the exact measure of his research had not yet been determined. This came to pass in Munich in consultation with Dr. Friedrich Hirth who, before his retirement, had been head of the Chinese Department at Columbia University. The invention of paper, an indisputable gift from China to the West, was fairly well known, even in sources other than the Chinese. The origins of gunpowder and the compass, while often attributed to the Chinese, were still so nebulous and vague even in Chinese sources that a scientific representation, if provable at all, would almost certainly mean many years’ work by many scholars. The invention of print¬ ing in China and its spread westward, while still to a large extent buried in Chinese sources and in archaeological material brought back from expeditions of Turkestan seemed, however, ready for the man who had the languages needed for the research and the patience and ability to put the scattered pieces of evidence together in a coherent account. At Dr. Hirth’s suggestion we postponed our return to America to spend the winter and spring in Berlin. Here my husband had a most rewarding time digging into Chinese sources and the treasure-house of archaeological material that had been brought back from Turkestan expeditions during the first decade of the century. Boxes never opened until this time were brought out for his inspection, and German scholars and scientists gave him freely of their time and knowledge. New clues at this time took him elsewhere on the continent. When research brought out the fact that paper containing print had been discovered in Egypt wrapped around ancient mummies, he went to a Berlin Arabist for more information. “Too bad you did not come a little earlier,” said Dr. Moritz, “for the man who knows more about this collection of papers than anyone in Europe left this office one hour ago to return to his home in Prague.” In less than a week we were in Prague, but by that time Professor Grohmann had left for Vienna. This time, however, the elusive scholar was fol¬ lowed immediately. As the collection of papers from Egypt was now in a Vienna museum it was a good piece of luck to have the collec¬ tion and the expert together in one place. Dr. Grohmann was most




The collection was taken out of the cases.


were made from the Arabic, and a chemist was called in to give expert opinion about the paper and ink. Happiness was ever one of Tom Carter’s most endearing char¬ acteristics, but never did it abound as during the months we had headquarters in Berlin and during our subsequent stay in France and England. In Paris he made personal contact for the first time with Paul Pelliot, the great Sinologist to whom he dedicated his book. During the winter’s research Mr. Carter had found evidence in Chinese sources that movable type as well as block printing had been invented by the Chinese, but up to this time he had seen no concrete evidence for the written statement.

During the conversation the

French scholar reached into a drawer in his desk, bringing out a small box. “If you are interested in the invention of movable type by the Chinese,

he said, “you will be interested in these specimens

of type. I found them on the floor of one of the caves at Tun-huang, and I have ascertained that they are considerably earlier than the Gutenberg invention.” The missing pieces to the story of the Chinese invention of printing and its spread westward were now falling into his hand almost daily. While in Europe Tom Carter accepted a call to join the faculty of the Department of Chinese at Columbia University, and in 1924 he was made its executive head. In teaching he found his fullest ex¬ pression. His own eagerness to learn, his attention to details, and his infectious enthusiasm were communicated to his students, a fact to which this writer can bear testimony. His classes became research laboratories where teacher and students alike contributed to the common aim. During his long illness in the spring of 1925 his stu¬ dents of Chinese civilization decided to conduct their own classes as seminars under such direction as he could give from his sickbed. The experiment was a great success and there was no falling off in attendance.

Several of his students who had started the semester

without a definite aim decided then to make the interpretation of the East to the West their life work. In addition to the scholarly incentive, the natural satisfaction that came to him in finding all his faculties used to their utmost capacity



in the research and in the writing of this book, there was present in Tom Carter’s work at this time another less obvious dynamic: an urgent desire to break down by intelligent knowledge the barriers that prevented understanding and appreciation between the East and the West. He had come to feel that there was as great a need for an interpretation of the East to the West as for the West to have inter¬ preters of its civilization in the East. It was this ideal that prompted him to accept the call from Columbia University. And it was this ideal, as much as scholarly ambitions and satisfactions, that made available in the West, for the first time, the history of the Chinese invention of printing, which became his parting gift to the life he loved so well. He was stricken with a fatal malady while the book was still in the press and passed away a few days after he had seen the completed volume in print. That the book filled a great need was made apparent in its en¬ thusiastic reception. It was placed on the League of Nations list of fifty books considered by the League’s cultural committee to have contributed most to essential world knowledge in 1925. When the first printing was exhausted in 1930, the publishers considered the possibility of bringing out a revised edition, but so little new material had actually appeared at that time that the project was abandoned. The new printing contained one correction on page 12, a few new titles provided by the late Dr. Berthold Laufer foi the bibliography, and a short biographical sketch of the author. When the book again went out of print in 195°* the Columbia University Press helpfully consented, at my request, to postpone another reprint until there had been time to investigate whether a revised edition could be made at this time. Talking over the problem with Dr. Carrington Goodrich, my husband’s successor in the Chinese Department at Columbia University, I found to my great satisfaction that he was willing to undertake the task. I wish to express here my deep appreciation for the spirit in which Dr. Goodrich entered on this by no means easy task. The additional material would, I am sure, have delighted Tom Carter. Probably it would have made him feel as he did when he defended his doctor’s thesis. “Was it hard?”



I asked him when it was all over.

'Why no,” he replied,


was a

most interesting experience. Think of getting all those fine minds working on my subject.” Except where later research has brought out facts that made re¬ vision necessary, the original text is left intact.

Throughout the

revision Dr. Goodrich has shown fine sensibilities in preserving Tom Carter s literary style, which indeed had no small share in the orig¬ inal success of the book. For this I am deeply grateful. I also wish to express my great appreciation for the unfailing courtesy and helpfulness I have received during these years from Columbia University Press, and for the understanding cooperation they have now shown in allowing the book to be transferred to a commercial publishing house to seek a wider distribution for this book so essential to Far Eastern as well as to world history. Dagny Carter

New York March, 1955

Part 1



Back of

the invention of printing lies the use of paper, which is the

most certain and the most complete of China’s inventions. While other nations may dispute with China the honor of those discoveries where China found only the germ, to be developed and made use¬ ful to mankind in the West, the manufacture of paper was sent forth from the Chinese dominions as a fully developed art. Paper of rags, paper of hemp, paper of various plant fibers, paper of cel¬ lulose, paper sized and loaded to improve its quality for writing, paper of various colors, writing paper, wrapping paper, even paper napkins and toilet paper1 *—all were in general use in China during the early centuries of our era. The paper, the secret of whose manu¬ facture was taught by Chinese prisoners to their Arab captors at Samarkand in the eighth century, and which in turn was passed on by Moorish subjects to their Spanish conquerors in the twelfth and thirteenth centuries, is in all essential particulars the paper that we use today. And even in our own times China has continued to furnish new developments in paper manufacture, both the so-called “India paper” and papier-mache having been introduced from China into the West during the nineteenth century.2 Though the invention of paper is carefully dated in the dynastic records as belonging to the year


105,3 the date is evidently chosen

rather arbitrarily, and this invention, like most inventions, was a gradual process. Up to the end of the Chou dynasty (256


through China’s classical period, writing was done with a bamboo pen, with ink of soot, or lampblack,4 upon slips of bamboo or wood. Wood was used largely for short messages, bamboo for longer writ* The notes to each chapter will be found at the end of the chapter. The bibliography and the list of characters follow the final chapter.




[Ch. i

ings and for books. The bamboo was cut into strips about nine inches long and wide enough for a single column of characters. The wood was sometimes in the same form, sometimes wider. The bam¬ boo strips, being stronger, could be perforated at one end and strung together, either with silken cords or with leather thongs, to form books. Both the wooden strips and those of bamboo are carefully described in books on antiquities, written in the early centuries of the Christian era. The abundance of wooden and bamboo slips excavated in Turkestan conforms exactly to the early descriptions. The invention of the writing brush of hair,5 attributed to the general Meng T’ien in the third century


worked a transforma¬

tion in writing materials. This transformation is indicated by two changes in the language. The word for chapter used after this time means “roll”; the word for writing materials becomes “bamboo and silk

instead of “bamboo and wood.” There is evidence that the

silk used for writing during the early part of the Han dynasty con¬ sisted of actual silk fabric.6 Letters on silk, dating possibly from Han times, have been found together with paper in a watchtower of a spur of the Great Wall. But as the dynastic records of the time state, “silk was too ex¬ pensive and bamboo too heavy.” The philosopher Mo Ti, when he traveled from state to state, carried with him many books in the cart tail.7 The emperor Ch’in Shih Huang set himself the task of going over daily a hundred and twenty pounds of state documents. Clearly a new writing material was needed. The first step was probably a sort of paper or near-paper made of raw silk.8 This is indicated by the character for paper, which has the silk radical showing material, and by the definition of that char¬ acter in the Shuo wen, a dictionary that was finished about the year A.D. IOO.9

The year


105 is usually set as the date of the invention of

paper, for in that year the invention was officially reported to the emperor by the eunuch Ts’ai Lun. Whether Ts’ai Lun was the real inventor or only the person in official position who became the patron of the invention (as Feng Tao did later with printing) is uncertain. In any case his name is indelibly connected with the in-




vention in the mind of the Chinese people. He has even been deified as the god of papermakers, and in the T’ang dynasty the mortar which Ts’ai Lun was supposed to have used for macerating his old rags and fish nets was brought with great ceremony from Hunan to the capital and placed in the imperial museum. The following is the account of the invention, as written by Fan Yeh in the fifth century in the official history of the Han dynasty, among the biog¬ raphies of famous eunuchs: During the period Chien-ch’u (a.d. 76-84), Ts’ai Lun was a eunuch.10 The emperor Ho, on coming to the throne (a.d. 89), know¬ ing that Ts’ai Lun was a man full of talent and zeal, appointed him a chung ch’ang shih.11 In this position he did not hesitate to bestow either praise or blame upon His Majesty. In the ninth year of the period Yung-yuan (a.d. 97) Ts’ai Lun be¬ came shang fang ling.12 Under his instruction workmen made, always with the best of materials, swords and arrows of various sorts, which were models to later generations. In ancient times writing was generally on bamboo or on pieces of silk, which were then called chih 13 But silk being expensive and bam¬ boo heavy, these two materials were not convenient. Then Ts’ai Lun thought of using tree bark, hemp, rags, and fish nets. In the first year of the Yiian-hsing period (a.d. 105) he made a report to the emperor on the process of papermaking, and received high praise for his ability. From this time paper has been in use everywhere and is called the “paper of Marquis Ts’ai.” 14

The biographical note goes on to tell how Ts’ai Lun became in¬ volved in intrigues between the empress and the grandmother of the emperor, as a consequence of which, in order to avoid appearing before judges to answer for statements that he had made, “he went home, took a bath, combed his hair, put on his best robes, and drank • >> 15 poison. Two statements in this quotation have received ample confirma¬ tion from discoveries along the Great Wall and in Turkestan. In March, 1931, while exploring a Han ruin on the Edsin-gol, not far from Kharakhoto,16 the Swedish archaeologist Folke Bergman dis¬ covered what is propably the oldest paper in the world. It was found along with a Chinese iron knife stuck in a leather sheath, a



[Ch. i

badly shriveled water sack of leather, a crossbow arrow with bronze head and reed shaft, many manuscripts on wood, silk rags (includ¬ ing a piece of polychrome silk), and an almost complete raincoat made of twisted grass strings.17 Lao Kan, who later made a report on this precious piece of paper,18 informs us that of the seventyeight manuscripts on wood the great majority were dated between the fifth and seventh years of Yung-yiian (a reign period covering the years


89-105). On the latest was written: “5th day of the

1st moon of the 10th year of Yung-yiian,” or February 24,



Mr. Lao agrees that, just because the last of the dated wooden slips bears a dated inscription, one cannot conclude that everything in the hoard was cached away in that year. Nevertheless, he surmises that about this time, possibly a few years later (whether before or after Ts’ai Lun’s historic announcement will never be known), the paper was manufactured and dispatched to this lonely spot in mod¬ ern Ninghsia province. Other pieces of paper of early times dis¬ covered in Turkestan date from about a century and a half after the announcement by Ts’ai Lun.19 The statement concerning the materials used has also been thor¬ oughly confirmed. Examination of paper from Turkestan, dating from the third to the eighth centuries of our era, shows that the materials used are the bark of the mulberry tree; hemp, both raw fibers and those which have been fabricated (fish nets, etc.); and various plant fibers, especially China grass (Boehmeria nivea), not in their raw form but taken from rags. The discovery of rag paper in Turkestan, while confirming the statement in the Chinese records, came as a surprise to many West¬ ern scholars. From the time of Marco Polo until some seventy years ago, all oriental paper had been known as “cotton paper,” and it had been supposed that rag paper was a German or Italian inven¬ tion of the fifteenth century. Wiesner and Karabacek in 1885-87 showed, as a result of microscopic analysis, that the large quantity of Egyptian paper which had at that time recently been brought to Vienna, and which dated from about


800 to 1388, was almost

all rag paper. A subsequent examination of the earliest European papers showed that they, too, were made in the main from rags.




The theory was then advanced and generally believed that the Arabs of Samarkand were the inventors of rag paper, having been driven to it by their inability to find in Central Asia the materials that had been used by the Chinese. In 1904, this theory suffered a rude shock. Dr. Stein had submitted to Dr. Wiesner of Vienna some of the paper he had found during his first expedition to Turkestan, and Dr. Wiesner, while finding in it no pure rag paper, did find paper in which rags were used as a surrogate, the main material being the bark of the paper mulberry. The theory was changed to suit the facts. The Arabs of Samarkand were no longer the first to have used rags in the production of paper, but the first to have produced paper solely of rags. Finally, in 1911, after Dr. Stein’s second ex¬ pedition, paper of the first years of the fourth century was laid before Dr. Wiesner and was found to be a pure rag paper! Rag paper, supposed until 1885 to have been invented in Europe in the fifteenth century, supposed until 1911 to have been invented by the Arabs of Samarkand in the eighth century, was carried back to the Chinese of the early fourth century, and the Chinese record, stating that rag paper was invented in China at the beginning of the second century, was substantially confirmed. The use of paper, so far superior to bamboo and silk as a writing material, made rapid headway. It was still, however, regarded as a cheap substitute.20 Extensive improvements in its manufacture were made by Tso Po,21 a younger contemporary of Ts’ai Lun. The rec¬ ords of the next centuries contain abundant references to the use of paper and to certain special fancy and beautiful papers that ap¬ peared from time to time. In Turkestan, at each point where excava¬ tions have been undertaken, the time when wooden stationery gave way to paper can be fairly accurately dated. By the time of the in¬ vention of block printing all of Chinese Turkestan, so far as ex¬ cavations show, was using paper.22 The use of paper in China proper had apparently become general much earlier. The papers found in Turkestan show a certain amount of prog¬ ress, especially in the art of loading and sizing to make writing more easy. The earliest papers are simply a net of rag fibers with no sizing. The first attempt to improve the paper so that it would



[Ch. i

absorb ink more readily consisted of giving the paper a coat of gypsum. Then followed the use of a glue or gelatine made from lichen. Next came the impregnation of the paper with raw dry starch flour. Finally this starch flour was mixed with a thin starch paste, or else the paste was used alone. Better methods of macera¬ tion also came into use that proved less destructive of the fibers and produced a stronger paper. All these improvements were perfected before the invention was passed on to the Arabs in the eighth cen¬ tury and before the first block printing in China began. So far as an invention can ever be said to be completed, it was a completed invention that was handed over to the Arabs at Samarkand. The papermaking taught by the Arabs to the Spaniards and Italians in the thirteenth century was almost exactly as they had learned it in the eighth. The paper used by the first printers of Europe differed very slightly from that used by the first Chinese block printers five centuries or more before.

NOTES 1. For descriptions of paper napkins and toilet paper in China, written by Arab travelers in the ninth century, see for paper napkins, Reinaud, 1845: 24, 38; and for toilet paper, Renaudot, 1718:17. See also Sauvaget, 1948:11, 49. 2. Laufer, 1927:73-74, holds that Chinese wallpapers were first introduced into Europe by Dutch traders at the end of the seventeenth century, and refers one to Sanborn, 1905. Reichwein, 1925:45-46, indicates that the Germans and French were probably the first, in the seventeenth century, to produce wall¬ papers from patterns brought from China by the missionaries. 3. On the date of the invention of paper, see comment by Goodrich, 1951: 145. Three years before a.d. 105, paper is mentioned in the biography of Empress Teng (a.d. 81-121). 4. Cf. Wang Chi-chen, 1930:122. There is a tradition that grew up in the I ang dynasty that during the Chou dynasty writing was done as a rule by cutting in the bamboo or wood with a knife. Chavannes, 1905, discusses in full this theory and the reasons why it cannot be held. 5. The development of the brush has been much discussed in recent years. Cf. the remarks of Yetts, 1929:1, 14-17; of Pelliot, 1930:374-78; and of Erkes, 194r.127-30. B seems clear from these discussions that the brush may already




have had a long history before the time of Meng T’ien. Archaeological proof is given by the evidence of brush writing on Yang-shao pottery (ca. 2000 b.c.), on jade, and on oracle bones of the late Shang period (ca. 1300-1028 b.c.). See Feifel, 1941:390-91. White, 1933:698, also 1934:32 and Plate XLI, reported the discovery of a bone brush handle over six inches long and a tile, “which could well have been the painter’s palette,” in Tomb 7 of a group of tombs of the Han state (450-230 b.c.) found at Chin-ts’un, Honan province, a few years ago. The excellence of their manufacture argues a long period of development. 6. Silk was used as a writing material in Mesopotamia in the early Moham¬ medan period before the Arabs there started to use papyrus rolls. For this purpose white silk was dipped in gum and polished with a shell. See Grohmann, 1924:59. Grohmann suggests the likelihood that this use of silk was derived from India, which seems quite possible; but as the silk both of India and of Mesopotamia was imported from China, it would seem likely that the art of preparing silk for a writing material both in India and in Mesopotamia went back originally to a Chinese origin. It is possible that before Chinese silk reached India and western Asia “there was wild silk in the Near East, similar to the tussah silk of India, as several fragments of this nature were excavated at Palmyra, one at Dura, and another in Egypt.” Day, 1950:108. While in China the use of silk as material for writing quickly gave way to paper, silk remained the usual material for painting for several centuries and has never been entirely displaced. 7. Cf. Duyvendak, 1947:314. In the Chuang-tzu it is recorded: “Hui Tzu was a man of many ideas. His works would fill five carts.” Cf. translation of H. A. Giles, 1926:450. 8. Early authority has it that some chih was otherwise made. Blue, 1948:22, cites a passage from Ssu-ma Ch’ien (writing about 100 b.c.), Shih chi 129/ib, which she translates: “[The region] west of the mountains abounds in timber, bamboo, [products made from] paper mulberry. . . .” In her note on the word {u thus translated she cites the gloss of Ssu-ma Cheng of the eighth century who explained that \u “is the name of a tree from the bark of which chih can be made”; also the earlier comment of Lu Chi (261303): “the people south of the Chiang (i.e., the Yangtze river) spin its bark to make pu cloth, and pound it to make chih ‘paper’.” Swann, 1950:420, skirts the problem by translating the last phrase “the f{U tree.” This early chih may have denoted a silk scroll. Hummel, 1941:74, refers to two uses of the word in this sense in a.d. 25 and in the years 76-84, and cites an article on the subject, written by Ma Heng in 1926. 9. Lao Kan, 1948, draws attention to the interpretation of the definition in the Shuo wen chieh tzu by Tuan Yii-ts’ai (1735-1815): “It is silk refuse beaten and soaked in water; a predecessor of paper undoubtedly, but not paper.” 10. Hsiao huang men; i.e., he worked in the yellow-gate palace as a eunuch. 11. An official permitted to go into parts of the palace forbidden to others.




12. An official charged with direction of the manufacture of furniture, household articles, etc. 13. Chih is now the ordinary word for paper. The definition of the word in the Shuo wen, finished about the time of Ts’ai Lun’s invention, would indi¬ cate that to that writer it meant a form of paper or near-paper made of silk. The passage under consideration would seem to indicate that the word had also been applied to the pieces of silk fabric used for writing. This word has the silk radical as indication of material. Later the same word with the cloth radical substituted for that of silk is frequently used, but it is the form with the silk radical that has survived and is in common use today. 14. Hou Han shu, chiian 108, biography of Ts’ai Lun. 15. For continuation of the biography of Ts’ai Lun, see translation in Blanchet, 1900:13-14, or Hunter, 1947:50-52. 16. Lao Kan calls the place Tsakhortei, south of the Bayan Bogdo Moun¬ tains. 17. Bergman, 1945:153. 18. Lao Kan, 1948:496-98, with an illustration of the paper fragment. 19. For a fuller description of this paper see Chapter 13. 20. Hummel, 1941:74, draws attention to a scholar named Ts’ui Yuan, who died thirty-seven years after paper was first announced to the throne, writing to a friend as follows, “I send you the works of the philosopher Hsu in ten scrolls—unable to afford a copy on silk, I am obliged to send you one on paper.”

21. Tso Po (T. Tzu-yi) was a native of Tung-lai (modern Yeh-hsien, Shantung province) and flourished at the end of the Han period, according to the Shantung t’ung chih, 1911 edition, 199/63. He was known as a fine callig¬ rapher. This biographical notice does not say that he made improvements, but that he acquired the skills of his master Ts’ai Lun. Consequently, Hsiao Tzu^ (

.494)5 writing to Wang Seng-ch’ien (246-85), ejaculated: “The paper

°, TZj'yi’ toSether Wlth his ink and brushes, is especially fine; none can reach their degree of excellence.” (The commentary of the Shantung provincial his¬ tory is lifted without acknowledgement from the writings of Chang Huai-kuan a scholar and calligrapher of the first half of the eighth century, preserved in the Shuo fu 92/4b.) Chang Huai-kuan (92/63) quotes Wei Tan (179-253) as remarkmg that to be a successful writer one should have the brush of Chang Chih (latter half of the second century a.d.), the paper of Tso Po, and the ink made by himself (1 e., Wei) See also Chao Ch’i (d. a.d. 2m), fu chueh lu 2/i4a> fragments of which have been collated by Chang Chu (b. 1781). 22. Writing on wood continued longest at Miran, a Tibetan fort, which ap¬ pears to have been particularly backward. Writing on wood continued at Miran parallel with the use of paper-down to the eighth or ninth century. Stein, 192!.348, 462. In most places in Turkestan it ended several centuries earlier.




that the Chinese word yin 1 today denotes both print and

seal is suggestive. A study of the history of the word sheds consider¬ able light on the origin of Chinese printing.

During the Han

dynasty yin meant to authenticate by the impression of a seal on clay. When clay impressions gave way some time about the fifth or sixth century of our era to inked impressions in red, the same word was used. When Taoist priests used as charms the impressions of wooden seals several inches square inscribed with the name of Laotzu or some other worthy, these larger seals were yin. When later the manifolding of Buddhist pictures and texts began, this block printing was yin. With the advent of every new invention, from that of movable type in the eleventh century to that of the Linotype in the twentieth, the same word has done duty, and the word yin which today still means seal, also signifies every form of printing, taken in the broadest sense. Back of the seal and the seal impression—in the time of the Chou dynasty (before 255


a practice that reminds one of the

tearing of the laundry check in the Chinese laundries of America.2 When a contract was made, it was written in duplicate on the two ends of a stick of bamboo. The bamboo was broken and one end retained by each party. The fitting of the broken ends was the authentication of the contract. In like manner, when the emperor bestowed a patent of nobility, the token of that patent was one half of a broken piece of jade—the other half being kept in the imperial possession.3 With the advent of the great emperor Ch’in Shih Huang (246209


the unifier of China and the builder of the Great Wall,

and with the more complex organization which then began, the




broken pieces of bamboo and jade gradually gave place to seals and seal impressions.4 The great seal of the conqueror, brought from the southern state of Ch’u by the minister Li Ssu, and engraved with eight characters, was for centuries the seal of empire, and its for¬ tunes figure both in history and in romance. The transition from the broken jade to the seal—from the prim¬ itive matching of broken edges to the more advanced and compli¬ cated matching of impression and die—was a natural one. But it may have been hastened by events that were taking place in an¬ other part of Asia. Just a hundred years before Ch’in Shih Huang’s conquests, Alexander the Great had conquered a part of India and had brought Greek culture to certain countries of Central Asia which were not so far removed from the expanding borders of China. In the land that lies between Alexander’s empire and that of China—the region now called Chinese Turkestan—Sir Aurel Stein found a collection of deeds, the seals upon which show the strange mingling of influences, Eastern and Western, that was going on during the Han dynasty, the dynasty that followed Ch’in Shih Huang. The documents, written on wood, are all closed, bound with cords, and sealed, the devices of the seal impressions being in some cases Chinese characters, in others elephants and Indian emblems, in still others heads of Zeus, Eros, and Medusa.6 It is of course far from certain that this Hellenistic influence had penetrated beyond Turkestan and into China—still less certain that it had penetrated as early as the reign of Ch’in Shih Huang. On the other hand it is not an impossibility.6 With the Han dynasty

(202 b.c.-a.d.

220) the use of seals grew

steadily more common, both for private and for imperial use. Seal cutting came to be a fine art, and for perfection of workmanship the seals of this time have never been surpassed.7 They were made of jade, gold, silver, copper, ivory, and rhinoceros horn.8 The seal impressions of the Han dynasty that have been found are in one respect quite different from those of later times. The seals of this period down to the fifth century were cut in a mold. When they were applied on a flat surface they had to be inked, gen¬ erally with vermilion. The characters came out white on a red

Macmillan Sf Co., Ltd.

Wooden Stationery of the Han Period. Asian


pi. 47.

(From Sir Aurel Stein, On Ancient Central

Copyright 1933, Macmillan & Co., Ltd., London.)

CO _


bo s * bo ^ X 3 £ Os U El d £ O s * vj> v’*U


; !^«siv J1

S\, v\ 4f*


' y

. I




— £.:,. ^s#> *^4d»ca^»**3»BKM«isa^^ iyfesar- *.-•*'• 4 kj^wssw****-^ 4 ■•



* %&-*'


i ■




_rfV**->»* *


»■: ' ■ ••*E*’-^^'» -«.- rwsv.,.. v-*»» x ■ *v*


« .# \a'»>4f\ i $

! IL,

/i CC\J

_* °



Museum fiir Voikerkunde.

Primitive Tibetan Gharri Prints.

(Charms 10.4 X 2.06 cm. each.)

Clay Containers for Tibetan Charm Prints. To find the charm, the clay container must be broken open.

Though these charms may not be earlier than the twelfth century

represent a survival of the most primitive form of block printing.


From Sangim Agiz

near Turfan. Museum fiir Voikerkunde.




f(ao mentions one cutting of the Tao te ching in 708, erected at I-chou, in Hopei province; another of 736 at Ch’eng-tu; and several others down to 881.7 Parallel with the early development of block printing, this sort of lithography 8 was also going on in Buddhist monasteries—developing until whole books were being produced. Especially famous is the huge collection of texts of the Buddhist canon cut during the years


600 to 1100 on stone steles, 7,145 in

number, which were later erected in grottoes and in a subterranean chamber.9 The manuscript chamber at Tun-huang that contained the earliest block printed book, the Diamond Sutra of 868, contained a copy of the very same book in the form of lithograph rubbing. The two copies, the one printed from wood, the other from stone, both date from the ninth century.10 The stone prints found at Tunhuang make it evident that already in the ninth century the practice had begun of preparing stones with the special purpose of taking rubbings from them, and that at least as early as the first books from blocks of wood (and probably earlier) both single sheets and rollbooks were thus being printed from specially prepared blocks of stone.11 However, it was in orthodox Confucian circles, and as an aid to the correct transmission of the Classics, that the stone inscription and the inked squeeze had their chief importance. Even after block printing began, the rubbing from stone was still the one official and orthodox method for the reproduction of standard texts.

It was

the union of these two processes, the Buddhist and Taoist block print (itself perhaps based on the earlier Taoist seal-charm) and the Confucian rubbing, that produced the great official block printing activity of Feng Tao’s time and instituted the era—during the tenth to the fourteenth centuries—when all of China’s great literature was printed. The important memorial of 932 by Feng Tao and Li Yii, that lay back of this awakening, began: During the Han dynasty Confucian scholars were honored and the Classics were cut in stone in three different scripts. In T’ang times also stone inscriptions containing the text of the Classics were made in the Imperial School. Our dynasty has too many other things to do and cannot undertake such a task as to have stone inscriptions cut and




erected. We have seen, however, men from Wu and Shu [Kiangsu and Szechuan] who sold books that were printed from blocks of wood. There were many different texts, but there were among them no ortho¬ dox Classics. If the Classics could be revised and thus cut in wood and published, it would be a very great boon to the study of literature.12

It is thus evident that when the Confucian Classics were cut in wood—the event that marked the beginning of large-scale block printing—those in charge of the work had no idea of printing. They thought they were continuing the ancient practice of cutting inscrip¬ tions, using wood instead of stone, after the analogy of certain Bud¬ dhist and Taoist prints that they had seen, for the sake of ease and economy. It was thus that the wooden block and its printed impres¬ sion developed naturally from the stone inscription and its rubbing. The Buddhist and Taoist prints—which had developed from charms and seals—gave the idea of cutting the inscription in reverse and gave also a new technique for taking the rubbing. The stone inscription gave the official precedent. Having thus been one of the influences that gave birth to wide¬ spread block printing, the use of rubbings did not cease, but con¬ tinued a parallel existence. Gradually during the tenth century, the century that showed the greatest activity in the development of all duplicating processes, the emphasis veered more and more from the inscription to the rubbing made from it. In the year 992 there is a record of the making of books which contained facsimiles of the autographs of the great men of the Tsin and the Wei dynasties, taken from a tomb that had recently been looted.13 Stone rubbings were thus the recognized method of preserving exact copies of beautiful calligraphy.14 When the wood blocks of these books became broken through constant use, they were mended with silver bars,15 the im¬ pression of which could often be detected in the rubbing. During the later years of the Sung period these lithograph books of 992 were treasured as great rarities. Throughout the Sung dynasty books from stone blocks con¬ tinued to be published.16 From China the art spread to Japan, and in 1315 a large collection of books was there printed by this process. The taking of rubbings still continues in China as a means of making




exact duplicates of ancient inscriptions, and there is no indication that the method has materially changed from the earliest times.

NOTES 1. The criticism of Moule, 1926:141, seems valid, namely that the practice of making rubbings of stone inscriptions as a forerunner of printing may be exaggerated. As Peake, 1935:12, citing Laufer and Yetts, points out, long before inscriptions, on stone there were inscriptions on bronze. This was done, wrote Laufer, “by means of the lost wax process, the characters being traced in the wax mould, and being either incised or raised in the bronze.” On moulds made for this purpose in Shang times (end of the second millennium before our era) see Karlbeck, 1935:39-60. To inscriptions on bronze should also be added those on pottery; cf. Chou Chao-hsiang, 1929:29-38. 2. The word used is mu-hsieh. The regular word for rubbing t’a did not come into use until the T’ang dynasty. The view is sometimes held that the word mu as used in the Han dynasty was the equivalent of t’a, but it is by no means certain. 3. Hou Han shu 5o/8b-9a, biography of Ts’ai Yung. For a note on the discovery of a fragment of one of these stones in 1929, see Gardner, 1938:594. There are a number of objections to this traditional view according to which the making of rubbings began in the second century. It seems more probable that the practice began not so very long before the date of the earliest rubbings found—perhaps in the sixth century. 5. This rubbing bears the date equivalent to 654 as the time when a certain person saw it, and the text is a poetical work composed and written by the emperor T’ai Tsung (627-49). Literary sources, however, indicate that ink squeezes were made a century and a half earlier, not only from stone slabs but also from wooden negatives of stone slabs. This is significant as it indicates that positives could be made directly from them—an important step in the prehistory of printing. See Pelliot, 1953:21-24. 6. According to the official history of the T’ang dynasty these officials were known as t’a-shu-shou. Des Rotours, 1947-48:1, 173, informs us that three of these officials were established in 649 at the College for the Development of Literature. In 731 a similar worker was given a post in the palace library (p. 194). The palace appointed six about 717 (p. 198). In 739 the College for the Exaltation of Literature had two such employees (II, p. 584). 7. Wang Chung-min, 1927:519-23. 8. Moule, 1926:141, criticizes the use of the term lithography here and elsewhere, but a more exact term is hard to come by. “Ink squeeze’’ or the



[Ch. 3

French word estampage also might be used to indicate the difference from the modern process of lithography. 9. Cf. Vaudescal, 1914:375-459. 10. Portions of another book of rubbings, cut and mounted in leaves, were found at Tun-huang. It consists of rubbings from the Hua-tu-ssu pei, of which the original was written by Ou-yang Hsiin (557-641). See Stein, 1921:11, 918, also IV, Plate 169. Some of the leaves of this book are in the Stein collec¬ tion in London and some in the Pelliot collection in Paris. 11. Literary evidence of books reproduced from stone blocks prepared espe¬ cially for the purpose goes back only to the tenth century. The clearest record is that of 992, quoted in note 14 below. There is, however, an earlier reference by Ch’u Po-hsiu (fl. 1270-87), quoted by Chou Mi (1232-1308), in the book entitled Yiin yen \uo yen lu (Ts’ung shu chi ch’eng edition No. 1553) 2/63: “The prince Hou-chou of the kingdom of Chiang-nan (961-75), of the Li fam¬ ily, ordered Hsii Hsiian (916—91) to take autographs of former dynasties which he possessed and make a collection of ancient and modern rubbings of them. These were engraved on stone and called Sheng-yuan t’ieh [after the period Sheng-yiian, 937—42]. But this puts the time before the Shun-hua period (990-94), and so we see in them the ancestor of the fa-t’ieh.” This, however, is most unlikely. Prince Hou-chou is Li Yii, born in 936. It is out of the question that he should have had a collection of autographs and ordered someone to make rubbings of them at the age of six or less. On this and other possible references to tenth century collections of auto¬ graphs engraved on stone, see Pelliot, 1953:93-104. The only evidence of books of rubbings in the ninth century is the actual discovery of these books at Tun-huang. 12. Ts'e fu yuan \uei 68o/3ob-3ia. 13. These volumes (there are ten of them) are known as fa t’ieh. Ou-yang Hsiu in Chi \u lu 4/9ab thus describes the production of the fa t’ieh of 992: “In the troubled times at the end of the T’ang dynasty, the Chao-ling, or imperial tomb of T ang T ai-tsung, was broken into by robbers, and the books and pictures that had been kept in it were torn from their rolls. Gold and jewels were taken and the books thrown away. Thus the autographs of great men of Tsin and Wei times came into the market. In the time of T’ai Tsung (976-98) these were bought up and put into ten books in order that they might be reproduced and passed on to posterity. These volumes were presented to the high ministers of state, and are the fa t’leh now in possession of various nobles and ministers.” According to Ts’ao Chao, in the work entitled Ko \u yao lun 3/1, these fa t’leh were published both in block print and in lithograph: “T’ai Tsung of the Sung dynasty searched out the autographs of men of former times and in the period Shun-hua (990-94) ordered the secretary Wang Chu to print them in facsimile in ten chuan. These were cut in blocks of jujube wood, and placed in the private cabinet of the emperor. ... In the third year of Shun-hua (992)




an edict was issued to cut these facsimiles in stone, and by use of Ch'en-hsint’ang paper and the ink of Li T’ing-kuei | /?. 937—75 ] to make rubbings. They were made in such a way that if you pass your hand over them the ink will not soil your hand.” After reviewing the evidence—including the statements of Wang Ying-lin in his Yu hai (ca. 1270) and of Liu Shih-ch’ang in 1299—Pelliot, 1953:105-20, considers that Ou-yang Hsiu and Ts’ao Chao were mistaken and that it is almost certain that the original engraving was on stone alone, not on wood. 14. That the use of the lithograph for preserving calligraphy had begun at an earlier date than this is indicated by the Tun-huang booklet described in note 10 above. 15. “Silver bar” is a tentative translation of Yin ting. 112-14.

Cf. Pelliot 1953:

16. The Ko chih ching yuan 39/6b-nb, devotes ten pages to a description of the lithographic texts produced in the Sung dynasty and three to counterfeits of these Sung lithographs. According to Julien, 1847:510, the Chih pu tsu chai ts’ung shu, compiled 1774 (2d ed., 1802), describes “all the ancient inscrip¬ tions and all the autographs of famous men that w'ere printed by this method between the years 1143 and 1242.” (This matter will be found in the above mentioned collectanea, section 10: it is taken from a two chiian work entitled Shih \’e p’u hsii, collected by Tseng Hung-fu, and originally printed in white on black in 1248.)


Art is

not the only expression of human genius which has been de¬

pendent for its greatest manifestations on strong religious feeling. It can be said with equal truth that every advance into new territory made by printing has had as its motive an expanding religion. In the whole long history of the advance of printing from its begin¬ nings in China down to the twentieth century, there is scarcely a language or a country where the first printing done has not been either from the sacred scriptures or from the sacred art of one of the world s three great missionary religions. China began by print¬ ing Buddhist pictures and texts.1 Japan had printed for six centuries and brought the printing of books to the highest degree of perfec¬ tion before the printing of anything but Buddhist sacred literature was attempted. The great mass of printed literature found in Central Asia continuing up to the time of the Mongol conquest is almost exclusively religious, consisting of Buddhist pictures and Buddhist books. The printing that was going on in Egypt through the time of the Crusades consists of verses from the Koran and of prayers. The block pi inters of Europe produced biblical pictures and the Poor Mans Bible, while Gutenberg printed the Bible itself. And in the nineteenth century the languages of Africa and the islands of the sea have been reduced to writing and to printed form almost wholly by missionaries, for the purpose of printing the scriptures. Even in China herself after the use of movable type had been almost forgotten, it was missionaries who reintroduced them to the land of their birth.2 If we expect, then, to find a strong religious impulse back of the invention of printing in China, we shall not be disappointed. The time when all sorts of experiments were being tried in various forms of reduplication—experiments that finally led the way to printing26




was the one strongly religious period in Chinese history. Under the powerful Han dynasty that ruled China for two centuries before and two centuries after Christ, men had not felt so strongly the need of religion. Reverence for the masters of the classical age just gone by seemed to be enough. True, there are records of Buddhism in China during the first century of our era, but so long as the united empire remained, the new religion made rather slow progress. About the beginning of the third century, however, the Han empire broke up and four hundred years of anarchy set in, sometimes com¬ pared to the Dark Ages in Europe, and caused by that same rest¬ lessness among the populations of Central Asia that spread such terror in Europe. For four centuries war was chronic—civil war and war with the northern barbarians. This age of anarchy may be roughly divided into the time of the Three Kingdoms, when three warring Chinese dynasties strove for the mastery; the Tsin dynasty, when China was again rather weakly united and fighting a losing battle against the barbarians on the north; and the period of division between North and South, when North China was in the hands of various Tatar dynasties. During this time literature went backward, and the settled, rather static culture of the Han times was broken up. It was no time for the conservative virtues of Confucian society. A religion that offered a way of escape from this sinful, distressed world had more chance. Buddhism steadily advanced throughout these four centuries. Wherever there was an especially beautiful spot or a location hallowed by some sacred memory, a temple or a pagoda was built, and the religious life, the life of retirement from the world, came to be the ideal of an ever-increasing multitude. A number of the pagodas of this period are still standing—among the oldest monuments we have of China’s Buddhism. The age of an¬ archy, especially its last century, was also an age of faith. With Buddhism came art. Not that all Chinese art is of Bud¬ dhist origin, as has sometimes been claimed. There was an art, of purely Chinese growth, that formed the foundation for the develop¬ ment of this and the succeeding age. But it was the new life that came in with Buddhism which touched that old art and made it great. All through the Dark Ages, while literature languished, art



[Ch. 4

grew. For the “barbarians” who ravaged China were not the rude hordes of Attila. They had become strong Buddhists and, as Bud¬ dhists, were the inheritors of that Greco-Indian art which had flour¬ ished in the wake of Alexander’s armies. Ku K’ai-chih, the father of Chinese painting, lived in the fourth century. Through the fifth and sixth centuries most of the little dynasties that strove for mastery have more names of artists recorded than they have years to their credit. The painters were in the Chinese South rather than in the Tatar North. Their art was Chinese. But it was the new religion, pouring in through the North and seen first in the sculptures of Northern Wei, that transfused it and gave it new life. Soon after the establishment of the T’ang dynasty, Chinese art entered upon its greatest, most creative period. With religion had come art. With religion and art came the impulse to print.

NOTES 1. It seems quite possible, as already pointed out, that this Buddhist activity was preceded by a practice among Taoist charm makers that was very closely akin to block printing. After the period of more or less primitive Buddhist printing, the next great step forward was the printing of the Confucian Classics by Feng Tao in 952, which marked a new stage in the art. Each of China’s three religions seems, therefore, to have had its part. But the greatest part, at least during the early centuries, was that of Buddhism. Taoist literature has suffered much destruction in the course of its history, which explains in large part why there is less evidence of the contribution of Taoism to the beginnings of block printing than there is of Buddhism. Besides being frequently under attack by the court during certain reigns of the T’ang dynasty, the Taoists suffered great losses in two conflagrations of Taoist books and woodblocks ordered by Mongol emperors during the third quarter of the thirteenth century (see Chapter 10). 2. It is possible that a critic may cavil over the importance assigned here to religion. (Cf. Peake, 1935 = 14-15, and Wu, 1936, 141-44.) It is true that the first secure reference to printing in Chinese literature is in the petition in 835 of an official in Szechuan province asking the court to order that there be no more printing of private calendars in the Yangtze valley, from Szechuan to Kiangsm We are informed, too, of the printing before 865 of two diction¬ aries, the Tang yun in five chiian and Yu p’ien in thirty chiian. (See Chapter 8 ) Nevertheless there is mounting evidence that without the religious impulse the development of printing might well have been measurably delayed.

Part II



Europe reckons

the date of the invention of printing from the time

when typography was invented, and considers block printing as merely an important step in preparation. The Far East reckons the invention of printing from the time when block printing began, and considers movable type as rather an unimportant later addition. This distinction lies in the difference between ideograph and alpha¬ bet. The writing of the languages of Europe is based on an alphabet: for them the invention of typography is the invention of printing. The writing of the languages of the Far East is based on some forty thousand separate symbols: for them, until the large wholesale print¬ ing of recent years, movable type have seldom been practical or economical. For any land, the invention of printing is the invention of that form of printing which transforms the education and culture of the nation. China invented movable type, Korea and Japan made great use of them—all glory to the courage of the inventors who applied typography to a language of forty thousand signs when it had not yet been applied to an alphabet. But the printing on which the renaissance of the Sung era was based, the printing which both in quality and quantity has always been pre-eminent in the Far East, is printing from wooden blocks. The invention of xylography, or block printing, is the truly significant form of the invention for China. Block printing in Europe was always a more or less rude art as it was at first in China, an art down among the common people, that won scant attention from scholars. When the finer work of Gutenberg appeared, the ruder art naturally came to an end. In China early block printing was equally rude. It was displaced, how31




ever, not by type but by a better form of block printing. Feng Tao, who a century or more after the beginning of block printing im¬ proved the art and applied it to new uses, is usually regarded by the Chinese as the inventor of printing, and holds much the same place in Chinese history that Gutenberg holds in that of Europe. From his day printing became a fine art. The books of the Sung dynasty have never been surpassed in printing skill. Chinese books printed from modern type cannot compare with them. In fact, one reason why movable type never succeeded in displacing the block book is Chinese love of calligraphy as a fine art. In the making of pictures, too, the wood engraver’s art has been carried to a very high degree of perfection, especially in Japan. The invention of printing from wooden blocks was therefore the invention of printing in China. It is the invention that by quantity production has largely transformed China s culture. It is the invention that in its quality has produced China’s finest books.

A necessary prerequisite for printing is ink. Scholars have long pointed out how large a part the discovery of an oily ink played in preparing the way for Gutenberg’s invention. In the same manner, the way was prepared for the invention of block printing in China by the use of an ink which is known in English as India ink, but is described more accurately by the French encre de Chine. The history of this ink, known as ch’i} is obscure but there are indications that it may have been known in classical times, possibly even in the Shang period (or end of the second millennium


according to Professor Wang Chi-chen.2 He quotes Hsiin Hsu (d. a.d.

289), editor of one of the Bamboo Books, to the effect that the

writing on the Chu shu chi nien, dating from 299


was in ink.

Popularly, however, the Chinese ink used both for writing and for printing

has been ascribed to Wei Tan, who died in



Although there have been many improvements and fancy inks de¬ scribed, especially by Sung dynasty writers, there has apparently been

Ch. 5]



little change since Wei Tan’s time in the main constituents of the ink which is ordinarily used. This ink is made by placing a number of well-lighted wicks in a vessel full of oil, while over this is placed a dome or funnel-like cover of iron. When this is well coated with lampblack, the lamp¬ black is brushed off and collected on paper. It is then well mixed in a mortar with a solution of gum or gluten, and, when reduced to the consistency of paste, is put into little molds. The best ink is produced from the burning of particular oils, but the common and cheaper kinds are produced from fir wood.5 This ink is sold in sticks or elongated cubes. To prepare it for writing, it is rubbed in water on a smooth ink stone. Chinese ink is excellent for printing from wooden blocks. It makes a clean neat impression and is peculiarly indelible—so in¬ delible in fact that on certain blocks of paper found in Central Asia, which have lain so long under water that they have become miner¬ alized, the writing is still clearly legible. The ink used in block printing, whether in China, in Central Asia, in Egypt,6 or in Europe is practically uniform. The makers of the primitive block prints of Europe were not so accustomed to the making of this sort of ink, and most of their work has faded into a sort of brown, but the essential elements are the same. It is too early as yet to determine whether this uniformity of ink indicates a line of connection or whether it indicates merely that block printers everywhere used the ink that would make a clear impression. On the other hand, Chinese ink is not satisfactory for taking impressions from metal. It stands in globules on the metal surface and makes a rough impression. The first typographers of Europe, faced with this problem, solved it by using an ink whose pigment was dissolved in oil—after the analogy of the early oil painters. China also experimented with printing from metal blocks, and in Korea printing with metal type was done on an extensive scale. It seems probable that there too the use of an oily ink for printing from metal must have been discovered, although no evidence of such use has yet been found. For use with wooden blocks—which




constituted the great bulk of all China s printing—Chinese ink was eminently satisfactory.

There is no indication that the method of block printing has greatly changed through its long history. A description of the art as it is now carried on will give some idea of what block printing in China means and has meant at least for the past thousand years— since the time of Feng Tao. The material used is generally pear or jujube wood. The wooden plate or block, of a thickness calculated to give it sufficient strength, is finely planed and squared to the shape and dimensions of two pages. The surface is then rubbed over with a paste or size, occa¬ sionally made from boiled rice, which renders it quite smooth and at the same time softens and otherwise prepares it for the reception of the characters. The future pages, which have been finely tran¬ scribed by a professional person on thin transparent paper, are delivered to the block cutter, who, while the above-mentioned ap¬ plication is still wet, unites them to the block so that they adhere, but in a reversed position, the thinness of the paper displaying the writ¬ ing perfectly through the back.

This paper being subsequently

rubbed off, a clear impression in ink of the reversed writing still remains on the wood. With his sharp graver the workman then cuts away with extraordinary neatness and despatch all that portion of the wooden surface which is not covered by the ink, leaving the characters in fairly high relief. Any slight error may be corrected, as in our woodcuts, by inserting small pieces of wood. But the process is on the whole so cheap and expeditious that it is generally easier to replane the block and cut it again; for this mode of taking the impression renders the thickness of the block an immaterial point. Strictly speaking, the press of China would be a misnomer, as no press whatever is used in their printing. The thin paper receives the impression with a gentle contact, and a harder pressure would break through it. The printer holds in his right hand two brushes

Ch. 5]



at the opposite extremities of the same handle; with one he inks the face of the characters, and, the paper being then laid on the block, he runs the dry brush over it so as to take the impression. This is done with such expedition that one man can take off a couple of thousand copies in a day. Sometimes the work is divided, one man inking the block, another taking the impression. The paper, being so thin and transparent, is printed on one side only and each printed sheet (consisting of two pages) is folded back, so as to bring the blank sides in inward contact. The fold is thus at the outer edge of the book and the sheets are stitched together at the other.7 This is the form of printing on which is based the development of culture in eastern Asia for the past thousand years. It is this printing that will be considered in the next chapters.

NOTES 1. What this ch’i was is uncertain. The character pictures drops of water falling from a tree, and means today the varnish made from the sap of the lacquer tree. With this varnish a pigment made from iron sulphate is often used, and it seems reasonable to suppose that some such material constituted China’s ancient ink. Such an ink would never have been satisfactory for block printing. 2. Wang Chi-chen, 1930:119-21. Professor Wang believes that varnish or ch’i writing meant in early times black-ink writing, and was later mis¬ interpreted (p. 122). 3. There was another form of ink—red ink—that might have been satis¬ factory for printing. This is made of red oxide of mercury or cinnabar. It was apparently used by the Chinese of Shang times (end of second millennium b.c.). See Britton, 1937:1-3, and 1940:7. It was anciently produced in Shu (modern Szechuan). Cf. Wang Yii-ch uan, 1951:25-26. Later it was produced in Hunan and it seems not to have been rare. Mr. Wang notes that Li Ssu (d. 208 b.c.), in his memorial of 237 b.c. to the prince of Chin, asserted that it was the chief material used for red paint. Subsequently it was much used by magicians of the Taoist cult. See Johnson, 1928:595 79 taking impressions from seals (see Chapter 2).

h *s

used for

4. “In the most ancient times a bamboo twig was dipped in lacquer for writing. In mid-ancient times there was an ink stone (mo-shih) from which




ink could be produced by rubbing. In the time of the Tsin and Wei dynasties, ink in blocks was first made. It was made from the smoke of lacquer and from lampblack produced by burning pine wood. So the people of the Tsin dynasty commonly used a concave stone for rubbing the ink stick and collecting the dissolved ink.” Tung t’ien ch’ing lu by Chao Hsi-ku (/?. first half of 13th century). (Unhappily this passage, quoted from Chen Yuan-lung’s Ko chih ching yuan 37/203, does not appear in the collection of reprints entitled Shuo fu, edition of 1647, 95. But it does appear in the Cho \eng lu 29/449 of T’ao Tsung-i [ca. 1320-99]. Dr. Fu Lo-shu provided this reference.) “In ancient times there were two forms of ink, one from lampblack of pine and one from ink stone.’ After the Tsin and Wei dynasties we hear no more of ink stone, as the making of ink from lampblack became general.” Chao Kuan-chih (Sung dynasty), Mo ching (Ts'ung shu chi ch’eng edition N°- 3525) Ia- Prof. Wang Chi-chen, 1930:124, explains the term mo shih or ink stone as one of three things, namely, coal, graphite, and petroleum.” The name Wei Tan as the inventor of ink is given on the authority of Lu Yu (first half of the fourteenth century). See Wang Chi-chen, 1930:114, 124. Actually he is the first ink maker of whom we have reliable record. 5. Cf. Davis, 1857:11, 180; and Julien and Champion, 1869:129-40. 6. Black ink of lampblack and red ink of cinnabar were both used in from the dynastic period down through Greek, Roman, and Byzantine times, as well as later. The usual ink for writing on papyrus was made much like the Chinese ink and was also kept in a dry condition. There is a curious parallel between the use of cinnabar for imperial decrees in the early Byzantine empire and in China. The restriction of the use of cinnabar to the emperor began in Constantinople about a.d. 470. 7. This description is taken in the main from Davis, 1857:11, 176-77. We have preferred to make use of this early description, as the method here described is less likely to be influenced by changes brought from the West.


The period

of the T’ang dynasty (618-906)—the period during

which Chinese printing had its birth—was one of the most glorious in the history of China. The four centuries of disunion and weak¬ ness—China’s Dark Ages—had been brought to an end some thirty years before the T’ang era commenced. Under the first emperors of the new dynasty, during the seventh century and the early part of the eighth, the ancient glory of the empire was revived and en¬ hanced. Not only China itself, but East Turkestan, Korea, and a large part of Indochina were at one time or another brought under the control of the court at Ch’ang-an, while armies were sent over the passes of the Himalayas into Kashmir against certain Indian states and over the T’ien Shan range into the region of Samarkand against the rising power of the Arabs. The early T’ang emperors of the century or more before Charlemagne did in China much the same work that Charlemagne did in Europe in restoring the old Empire on a new basis and bringing to an end the long era of chaos and disorder. But the chaos of China’s Dark Ages had never been so complete as that of Europe, and classical civilization was first re¬ stored, then surpassed, far more quickly than in the Western world. The early emperors of the T’ang dynasty were great patrons of literature, of art, and of religion, and ruled over a people whose mental vision was rapidly expanding. Under T’ai Tsung (627-49), a library was erected at the capital which contained some fifty-four thousand rolls.1 At the same time, China’s attainment in the domain of painting was rapidly approaching its high-water mark. For impartiality in religious toleration, T’ai Tsung and his im¬ mediate followers have seldom been surpassed in history. While they themselves leaned toward Taoism and considered their family 37




to be of the lineage of Lao-tzu, they were liberal patrons of Confucian scholarship and welcomed with open hand every foreign faith. Within the space of thirty years, in the early part of the seventh century, the court at Ch’ang-an had the opportunity to wel¬ come the first Christian missionaries, to give refuge to the deposed king of Persia and his Mazdean priests, and to do honor to Hsiiantsang, the greatest of all the apostles of Chinese Buddhism, who returned from India to give new impetus to the Buddhist faith. All received the heartiest welcome. All propagated their respective faiths with the emperor’s favor and help. Contact with men of many lands and of varied opinions produced an alertness, a renewing of youth in the land, such as China had never before known. This Augustan age lasted for more than a century. It culmi¬ nated in the reign of Ming Huang (712-56) in whose time the Hanlin Academy was founded, and about whose court gathered such men as Li Po and Tu Fu, Wu Tao-tzu, and Wang Wei, the greatest poets and the greatest artists whom China in all her long history has known. During this golden age of Chinese genius, a great variety of devices was being evolved in the Buddhist monasteries and else¬ where for the reduplication of sacred books and texts—an activity that reached its climax in block printing some time before the end of the “golden age.” One of the earliest indications of the multiplication of illustra¬ tions in the East comes from the great Chinese Buddhist pilgrim I-ching (635-713). After a long sojourn in India (673-85), he spent several years translating Sanskrit texts on the island of Sumatra, whence in 692 he sent to China his report. One sentence of the report runs as follows: “The priests and the laymen in India make Kaityas or images with earth, or impress the Buddha’s image on silk or paper, and worship it with offerings wherever they go.” 2 It is puzzling to find I-ching applying this practice to India, where there was silk but where paper was rare. For China and its nearest neighbors, however, it seems entirely reasonable.3 This activity in devising methods of multiplication can best be studied from the finds of Tun-huang and those of Turfan, the two




places where the manuscript records of early Buddhism on the borders of China have been preserved.

(See Chapters 8 and 14.)

Here are found not only rubbings from stone inscriptions, but also stencils and pounces, printed textiles, seals and seal impressions, and a great profusion of little stamped figures of Buddha, all of which led the way directly to the art of the block printer. The rubbing from stone was in the main the Confucian prepara¬ tion for printing. But discoveries at Tun-huang show that the Bud¬ dhists used the device, too, and by means of it printed one of their favorite scriptures, the Diamond Sutra. (See Chapter 3.) The stencil or pounce was a means of reduplication of which the Buddhist monasteries were especially fond. Several of these paper stencils have been found, with large heads of Buddha first drawn with a brush, then outlined with needle pricks like a modern em¬ broidery transfer pattern.4

Among the finds are also stenciled pic¬

tures—on paper, on silk, and on plastered walls. Printed textiles0 appear in considerable number at Tun-huang. These are sometimes in two colors, sometimes in several. The de¬ signs are all conventional and nonreligious, an entire contrast to all other early printing and pre-printing in the Far East. Conven¬ tionalized animal designs—horses, deer, and ducks—are popular. There is also one example of design-printing on paper.6

It looks

like heavy modern wallpaper, with a dark blue geometric design. Small stamped figures of Buddha mark the transition from the seal impression to the woodcut.

Thousands upon thousands of

these stamped impressions have been found at Tun-huang, at Turfan, and at other places in Turkestan. Sometimes they appear at the head of each column of a manuscript. Sometimes great rolls are filled with them—one such roll in the British Museum is seventeen feet long and contains four hundred and sixty-eight impressions of the same stamp.

The only difference between these Buddha

figures and true woodcuts, other than the primitive workmanship shown, is that the impressions are very small,7 and hence were evi¬ dently made by hand pressure like the impressions from seals. The stamps found have handles for this purpose.8 When the idea oc¬ curred to some inventive genius to turn his stamp upside down,




lay the paper on it, and rub it with a brush, the way was open for making impressions of any size desired, and the way was open also for such improvement of technique as made the new invention a force in the advancement of civilization. But first it seems to have brought about only the making of better Buddha figures. One roll at London, though similar in many respects to the others, was evi¬ dently made not by stamping but by rubbing, for it shows much larger and better Buddha impressions.9 A perfected woodcut in the Louvre shows a still further advance—a number of Buddha figures in concentric circles of varying form, and all made from one block.30 Such are some of the steps—rubbing from stone, printed silk, stencil, seal, and stamp—that were leading at the same time toward the block print. All these objects have been found in Buddhist mon¬ asteries, and back of all, or most of them, lies that duplicating im¬ pulse that has always been a characteristic of Buddhism. That these actual objects found at Tun-huang and Turfan are earlier than the first block books is by no means certain. None bears clear indication of date except one stone rubbing and one stamp.11 But there is every indication that those which are not themselves earlier than the first block printing at least represent survivals of earlier and more primitive processes. The exact date at which true block printing began is shrouded in mystery. A supposed reference to printing as having taken place under the emperor Wen in 594, before the beginning of the T’ang dynasty—a statement that has found its way into almost everything that has been written in European languages on the subject of Chinese printing—is apparently based on an error by a Chinese writer of the sixteenth century.12 At this point it is necessary to mention that one fragment of paper, found near the then Chinese frontier, which bears a date equivalent to


594, has recently been reported as a printed item.

Discovered by Sir Aurel Stein during his third expedition to Central Asia in the years 1913-16 amongst the ruins of a Buddhist temple at the village of Toyuk (or Toyukh) in the neighborhood of Kara Khoja,13 it was turned over for study—after World War I—to Pro¬ fessor Henri Maspero, along with all other documents on wood and

Rubbing Being Taken from a Stone Inscription,

Museum filr Votkerkunde.

Stamps such as this mark the transition between the seal and the block print. (Height 6 cm.)

Metal Stamp for Making Figures of Buddha.

Museum fiir Volkerkunde.

Such rolls and fragments have been found in great quantities in various parts of Turkestan by British, French, German, and Japanese expeditions. (15.5 X 22 cm.) Fragment of a Roll of Thin Paper with Stamped Buddhas.




paper. Unfortunately for the world of scholarship, Maspero’s manu¬ script of some 600 pages, completed in 1936 and sent to London that same year, has only just been published.14 But a few years ago Dr. Bruno Schindler was entrusted with the preparation of a resume of Maspero’s findings and announced that this was a poster, printed in Chinese, “complete at top and bottom, but cut on right and left side. . . . The text reads (in translation): \ . . 34th year yen-ch’ang (= a.d. 594), year chia-yin. There is a vicious dog in the house. Passers-by to take care.’ ”15 This astounding information now ap¬ pears to be in error. On the authority of Dr. Harold James Plenderleith, Keeper, Department of Research Laboratory, The British Mu¬ seum, who has examined it, the document shows no indication of printing.10 Dr. Schindler too has withdrawn his earlier assertion, and considers that Maspero made a mistake.1' The difficulty of dating the beginning of block printing is en¬ hanced by the fact that the evolution of the art was so gradual as to be almost imperceptible. The earliest well-defined block print ex¬ tant dates from 770 and comes from Japan. The earliest printed book comes from China and is dated 868. But that printed book is a highly developed product. It is evident that the feverish activity in devising new ways of reduplication, which was going on in the Buddhist monasteries and elsewhere before this time, must have culminated in some sort of block printing before 770, and long enough before that date to have been by that time carried across to Japan. Perhaps the nearest approach to an approximate date that can be given would be the reign of Ming Huang (712-56), the time when China’s national greatness and China’s cultural achievement reached their height. The reign of Ming Huang ended in a disastrous revolution. The glories of the T’ang dynasty from that time began to fade. The policy of perfect toleration for all religious faiths that marked the reigns of T’ai Tsung and Ming Huang was abandoned, and in its stead there grew up a policy of persecution of foreign faiths, in¬ cluding Buddhism. This persecution culminated in the famous edict of 845, by which 4,600 Buddhist temples were destroyed and 260,500 Buddhist monks and nuns forced to return to lay life.18 It is




owing to this destruction of temples, as well as to the civil wars of the last century of the T’ang dynasty, that most of the great works of art of the T’ang period have perished. It is doubtless due to the same cause that no Chinese printing earlier than the Diamond Sutra of 868 has survived, and that for the earliest extant block prints it is necessary to turn to Japan.

NOTES i. In the first edition Carter, who gave no source, wrote that under T’ai Tsung a library was erected containing 200,000 volumes. This figure appar¬ ently comes from the T’ang hui yao 64/143 where in a discussion of the Hung wen \uan (College for the Development of Literature) it is reported that in the early years of his reign there were deposited in the palace books totaling over 200,000 chiian. In spite of the immense authority of this work, completed in 961, one may question the accuracy of the figure. Does it include duplicates? If so, no more need be said. If, however, it means distinct works it seems too large by several fold. According to the Sui bibliography (Sui shu 35/27^ there were in this period (a.d. 590-618) 3,127 works (pu) in 36,708 chiian, lost and duplicate works not counted. It must be mentioned here that the T’ang bibliography in the New History (Hsin T’ang shu 57/23) gives a figure for the beginning of the Sui of 370,000 chiian. Surely some scribe slipped at this point! The second emperor of the Sui set about having fifty duplicate sets made. (Sui shu 32/6b.) He had most of this library taken to his capital at Yangchow, where they were lost in the destruction of the palace in 618. When the remainder (8,000 plus) were started north under the care of Sung Tsun-kuei all but 10 or 20 per cent were lost in a shipwreck. (Sui shu 32/ja, Chiu T’ang shu 47/46^ Hsin T’ang shu 37/23.) At this point one must insert a word about Buddhist literature. Accord¬ ing to Fa-lin (572-640), Pien cheng lun 3, forty-six collections of sutra and sastra in 132,086 rolls were copied, and 3,853 old copies of sacred books were repaired during the reign of Sui Kao-tsu (581-604). Wright, 1951:35-36, draws attention to a Confucian lament, issued after the edict of 581 which ordered the copying of Buddhist texts at the expense of the state, “Among the people Buddhist scriptures are more numerous than the Six Classics by several thousand fold.” (Sui shu 35/363.) At the beginning of the T’ang (621) the imperial army subdued Wang Shih-ch’ung, who had seized Loyang together with its palace treasures, which included a library of over 80,000 chiian, some of them duplicates, some in¬ complete texts. These were transported by water to the capital at Ch’ang-an.




There were some losses en route, but these constituted the library of T’ang T’ai Tsung. Cf. Chiu T’ang shu 47746b and des Rotours, 1947-48:1, 191. An immediate effort was made to strengthen this collection by copying other books in private collections and by purchase, but even at the height of the T’ang (718-41) the figures reported are 53,915 chiian by pre-T’ang authors and 28,469 chiian by T’ang authors, a total of 82,384 chiian. (Hsin T’ang shu 57/ib.) The T’ang hui yao 64720b reports that in 721 there were 81,990 chiian. (Correct des Rotours, idem, who publishes the figure 80,990. His figure of 21,748 chiian of books in the field of philosophy should also be cor¬ rected to 21,548.) In another library at Loyang(P) there were in the year 744 books amounting to 54,574 chiian. The Chiu T’ang shu 47746b states that in the two capitals, Loyang and Ch’ang-an, there were 125,960 chiian. In the years 715-21 there was compiled a catalogue in 200 chiian entitled Ch’iin shu ssu pu lu which was revised a little later and appeared under a new title, Ku chin shu lu edited by Wu Ching (name given by T’ang hui yao 36 as Wu Chao), an editor of the earlier work. This listed 3,064 works in 51,852 chiian. (See Chiu T’ang shu 46/22 and Yii hai 52725b.) His separate catalogue of Buddhist and Taoist works entitled K’ai-yiian nei wai ching lu had over 9,500 chiian. (Hsin T’ang shu 59/i2b.) As des Rotours notes, the figures given are sometimes confusing, but they do seem to show that 200,000 volumes for T’ai Tsung’s library are too many. We have taken instead the number of chiian reported in the first half of the eighth century for pre-T’ang books. This lengthy digression points up the immense literary culture, even “bookishness,” of the Chinese people and hence the very practical need for an easier, quicker method of reproducing books. 2. Takakusu, 1896:150. 3. Note the discussion of this problem in Pelliot, 1953:14-19 and the note ibid.: 15, supplied by Professor Demieville. Liebenthal, 1947a: Plates 2-8, has drawn attention to many stamped bricks carrying inscriptions in Sanskrit, some negative, some positive. These seem to point to a kind of proto-printing in burnt clay. The bricks came from pagodas in Yunnan which Liebenthal be¬ lieves were constructed in the ninth century. (Personal communication to L.C.G.; cf. also Liebenthal, 1947b: 10-12.) 4. See Stein, 1921: index, “Stencil,” and Plates 94 and 113. 5. See Chapter 20. Also Stein, 1921: index, “Silk, printed,” and Plates 122 ff. 6. This forms the base or frame on which one of the block prints in the British Museum is pasted. The pattern is of ellipses forming a net, printed in dark blue on a light blue ground. There are two fragments, each about six inches by two. 7. The impressions on the rolls in the British Museum range from 1.5 to 2.8 inches in height and from 1.2 to 1.8 inches in width. Those in Paris and Berlin are approximately the same.




8. A wooden stamp found by Pelliot at Kutcha in Eastern Turkestan dates—according to the deposit in which it was found—from not later than 800. That the use of these stamps had spread as far west as Kutcha by 800 indi¬ cates a very early date for China itself. Metal stamps of uncertain date have been found at Turfan. 9. Size 4 x 3.4 inches. Only the bare outline is printed. Details are filled in by hand in colors. The workmanship of this sheet of heads bears a striking resemblance to the most primitive European block prints. 10. Size 13 x 20 inches. 11. This rubbing dates from the reign of T’ai Tsung (627-49); see Chap¬ ter 3. The stamp dates from before 800; see note 8 above. 12. Julien, 1847:505-7, was the first to introduce to European readers the view that printing was carried on in China in the year 593 (actually 594; see below), and from Julien the statement found its way into the Eleventh Edition of the Encyclopaedia Britannica and most histories of China in Western lan¬ guages. (Two others may also be held responsible for this assertion: Alexander Wylie in his Notes on Chinese Literature [Shanghai, 1867], p. xviii, and in his essay “Chinese Language and Literature,” reprinted in Chinese Researches [Shanghai, 1897], p. 237; and The Archimandrite Palladius in his ChineseRussian Dictionary, Vol. I, p. 264. See Laufer, 1927:72.) The origin of this theory is interesting. Julien quotes it from the Ko chih ching yuan, published in 1735. The statement in this encyclopedia is quoted from Lu Shen (1477-1544) and from the book Pi ts’ung. Lu Shen’s statement, contained in his book Yen hsien lu fol. 2ab, is: “Under the Emperor Wen of the Sui dynasty, in the thirteenth year of K’ai-huang, the eighth day of the twelfth month [January 5, 594], on orders from the emperor, all neglected hsiang [the word means either images or pictures] and scattered ching [Classic texts or sutras] were carved and composed. This is the beginning of the print¬ ing of books. It was thus earlier than Feng Ying-wang [i.e., Feng Tao, 881954].” This reference of printing to Wen’s reign is clear and explicit and indicates that the theory went back as far as the sixteenth century. The state¬ ment in the book Pi tsung (full title of the book is Shao shih shan fang pi tsung), by Hu Ying-lin (1551-1602), is apparently based on that of Lu Shen. It reads simply: Block printing had its birth at the beginning of the Sui dynasty, it expanded greatly under the T’ang, took a leap forward under the Five Dynasties and finally came to its fullest development under the dynasty of Sung.” Against these two statements is the weight of the older Chinese tradition (the very form of Lu Shen’s statement shows that he is propounding somet ing contrary to the general opinion), and also the explicit authority of at least three ^prominent writers of the Sung dynasty, whose statements follow: Under the Tang dynasty block printing, though carried on, was not fully developed. Under F£ng Ying-wang [FSng Tao] first the Classics and then all the ancient canonical works were printed.” Shen Kua, Mina ch’i t>i t’an 18/9. 5 r

Ch. 6]



“Before the T’ang dynasty all books were manuscripts, the art of printing not being in existence. . . . According to popular report the cutting of blocks and printing of books from them was commenced by Feng Tao. This is not the fact. . . . Printing certainly existed in the T’ang dynasty, but I apprehend it was not equal in workmanship to the present.” Yeh Meng-te (1077-1148), as quoted by Ma Tuan-lin in Win hsien t’ung f^’ao. Translation by Meadows, in Curzon, i860. “There was no printing before the T’ang dynasty. Inked blocks were first used at I-chou at the end of the T’ang dynasty.” Chu Yu, 1 chiao liao tsa chi 2/61. In the face of such conflicting evidence it is necessary to discover where Lu Shen got his information, which has so remarkably dominated European writings on the subject. There is apparently nothing about printing in the annals of the Sui dynasty. In the Buddhist Tripita\a, however, in the volume by Fei Ch’ang-fang entitled Li tai san pao chi 12/666 stands the passage from which Lu Shen’s statement is a word for word (though abbreviated) quotation. This book was written in 597, only three years after the event related. The last two sentences of Lu Shen’s statement (“This is the beginning . . .”) are not quotation but Lu Shen’s comment. A critical examination of the whole passage, without this gloss and in the context, leaves little doubt that printing is not referred to at all, the true interpretation being that damaged images were recarved and that scattered sutras were collected. This interpretation of the passage was first proposed by Yuan Tung (1697-1761) in the book Shu yin ts’ung shuo, and accepted by the Japanese investigator, Kamezo, 1909:3. Arthur Waley, 1919:413-16, independently reached the same conclusion. Yeh Te-hui in Shu lin ch'ing hua 1/19-20 gives a full discussion of Lu Shen’s inter¬ pretation of Fei Ch’ang-fang’s statement, with quotations from Chinese and Japanese authorities. His conclusion is that Lu Shen’s interpretation is incor¬ rect and that there is no reference to printing. Pelliot, 1953:1-10, agrees. 13. Stein, 1928:609-19. 14. Maspero, 1953:167. 15. Schindler, 1949:238-39 and Fig. 4. 16. Personal communication to L.C.G., dated May 16, 1953. In Dr. Penderleith’s words, “the evidence is entirely in favour of calligraphy.” 17. Schindler, 1952-53:222-23. His communication is dated June 15, 1953. 18. Certain of the later emperors of the T’ang dynasty were completely under the influence of Taoist superstition, and to this, together with other in¬ fluences such as the urgency of the economic situation, was due the persecution of Buddhism that lasted from 845 to 859. In 859 Buddhism was restored to its former position. It is well to note that the persecution was not enforced to any degree during the later years. The Ch’ing-lung monastery was restored in Ch’ang-an in 846 and the Fo-kuang monastery at Wu-t’ai shan in 857. Add to this the fact that certain monasteries in distant parts of the empire, such as Szechuan and Chekiang, completely escaped damage.


For a century

and a half before the making of the first block-

printed charms, Japan had been undergoing a process of complete transformation under the influence of China.1 It was a period similar to that which Japan passed through during the latter half of the nineteenth century, except that China was the model instead of the West. A steady succession of Buddhist missionaries from China poured into Japan, and a steady succession of Japanese students went to China for study and on their return brought about sweeping changes in the customs of their native land, bringing Japan gradu¬ ally abreast of what was then the world’s most cultured country. In 701 the annual celebration in honor of Confucius began, and in 708 the first mint was established for the making of coins in Japan. 735 a Chinese scholar “ became head of the newly established uni¬ versity at Nara, Japan s new capital, which was seeking in every way to mold itself after the pattern of the Chinese capital at Ch’angan. In the same year Kibi no mabi (d. 776) returned from Ch’ang-an after nineteen years of study and, entering into the service of the government, introduced information about the Chinese calendar and laws and other Chinese customs. To him is ascribed (doubt¬ fully) the invention of katakana, the Japanese syllabary or script. He was the tutor of the Empress Shotoku, by whose order the first recorded block printing was done. A recent Japanese writer has given the following account of the zeal with which Japan was at this time adopting Chinese ways and culture: 46

Ch. 7J



During the eighth and ninth centuries there was scarcely anything good in Si-an-fu, the great T’ang capital, that was not introduced into Japan or copied by the Japanese in their capital at Nara sooner or later. If the court buildings at Si-an-fu were painted red, so were those at Nara. If a temple was built and supported by the Chinese government in each province, so must it be in Japan. If the birthday of the Chinese emperor was observed as a national holiday, so was it here. If the nobles and the upper class in the Chinese capital played football, it was soon imitated by the Japanese aristocracy in Nara. . . . We can trace all this back to the Chinese origin of Japanese Bud¬ dhism.3

In Japan as in China, block printing was preceded by the use of seals. As early as the year 629, reference is made in the Nihongi to the imperial seal.4 In 704 official seals for the provinces were estab¬ lished, and it was stated that they were to be two sun square (a little more than two inches).5 In 739 a seal of the same size was granted to the Ise shrine. These seals without doubt followed the fashion that was already in use in China and were used for making impres¬ sions with ink. That some of them, at least, were made of wood is indicated by the statement in the Nihongi that in 692 the office of the Shinto cult gave a wooden seal to the empress. Japan, the country that was never conquered until the last decade, is remarkable for the careful way in which ancient antiquities have been preserved. This is particularly true of the town of Nara, where the capital was established from 710 to 784, and where a large vari¬ ety of objects from this Nara period have been kept. Among the precious objects preserved at Nara are a number of pieces of printed silk fabric which were apparently made by the use of wooden blocks. The patterns include plants, flowers, willow trees, pheasants, small birds, and butterflies. Two of the pieces of silk have the date printed into the design—dates corresponding to the years 734 and 740.6 Printed textiles, or surigoromo, are mentioned in the Sho\u Nihongi under the date of 743. Armor belts of leather, with designs in blue, red, and purple dye printed upon the leather, were produced at various times in the provinces of Hisen and Higo in the southern island of Kyushu, and some of them have been preserved. One is dated the eighth month




of the twelfth year of the period Tempyo, which corresponds to 740. It is even more close to being a true block print than are the textiles, for the printing includes not only design but also a picture of the divinity Fudo and a number of words in Chinese and San¬ skrit as well as the date.7 During the whole of the Nara period (710-84) the control of the Buddhist hierarchy over the affairs of the empire was very strong. The resources of the state were drained for the casting in 732 of the forty-nine ton bell—the fourth in size in the world—and for the erection in the years 735-49 of the great bronze statue of Buddha at Nara, weighing over five hundred and fifty tons and covered with fifty pounds of gold. The priest Gembo, who returned from China in 735 after an eighteen years’ stay,8 and who brought back with him five thousand Buddhist books and many holy images, had a large share in managing the affairs of state until his death in 746. But it was under the Empress Shotoku, who reigned, with in¬ terruptions, from 748 to 769, that priestly control reached its climax. This empress, remembering the terrible smallpox epidemic of 735-37, kept a hundred sixteen priests attached to her court for the driving out of disease demons, in addition to those employed for other purposes. Dokyo, the head of the Buddhist priesthood, was her chief physician and adviser and had a controlling voice in all state decisions. He was emperor in everything but name, was even given several of the titles usually reserved for the emperor, and was lodged in the palace. To the zeal for Buddhism of the Empress Shotoku, the world owes its first certain and clearly attested record of printing with copper blocks upon paper.9 The empress ordered the printing of one million charms to be placed in a million tiny wooden pagodas, and some time before the year 770 the work was finished and the pagodas and the charms distributed. This event, so important in the history of the world, rests fortunately on as clear evidence as any event in early Japanese history. It is described both in the dynas¬ tic annals and in the records for the temple where many of the prints were deposited. More than that, a number of the original prints are still extant.




The account in the official history, the Sho\u Nihongi, is as follows: In the fourth month of the year 770,10 after the eight years of civil war had been brought to an end, the empress made a vow and ordered the production of one million three-story pagodas, four and a half inches high and three and a half inches in diameter at the base. Within each of the pagodas was placed a single copy of one of the four dharanl [here follow the names of the four charms]. When this work was finished, the pagodas were distributed among various temples.

The record in one of the temples is more explicit with regard to the means by which the charms were made: In the year 767 there were built two small halls for pagodas on the east and west sides of the temple. . . . There were made one million pagodas, which were divided among the following ten temples [names of the temples]. In each was preserved a charm (dharani) from the Mu-\u Jd-\o Sutra in block print.11

Not only do we have these two clear contemporary accounts of the printing of a million charms but we also have the charms them¬ selves. A number of the original impressions are preserved in the Horyu-ji, a monastery in the province of Yamato, together with at least 109 of the little pagodas in which they were contained. Nine public libraries and museums and several private collections in the United States have these charms.12 One museum in Canada also has a charm, together with its reliquary or pagoda, and two private col¬ lectors likewise have reliquaries. The charms are about eighteen inches long by two wide. Each one contains about thirty columns of five characters each. They are not all alike, as four different charms were printed.13 Two different kinds of paper were used, one thick and of a woolly texture, the other thinner and harder, with a smooth surface, which did not absorb the ink quite so readily. All the charms, on both kinds of paper, are brown with age. The text of these earliest block prints and of the whole Sutra from which they are taken indicates clearly the incentive that was back of their production, and sheds light on the powerful impulse




that Buddhism gave to early printing. This Buddhist classic 14 con¬ sists of six sections, each of which in turn contains a narrative portion and a charm, the narrative portion indicating the use of the charm. When, sometime prior to 704, the Sutra was translated into Chinese by Mi-t’o-hsien—sixty years before the printing of the charms in Japan—only the narrative portions were translated.

The charms

were merely transliterated, the Sanskrit sounds being represented as nearly as possible by Chinese characters. It is these Sanskrit charms in Chinese characters that were printed and rolled up and placed in the wooden pagodas. A small section from the narrative portion of the Sutra, which forms as it were the introduction to the charms, is enough to indicate how this printing naturally fitted into the Bud¬ dhist scheme of salvation: A Brahmin who was sick went to visit a seer in a garden. The seer said, “You must die in seven days.” So he went to Buddha, plead¬ ing that Buddha would save him, and offering to become his disciple. Buddha said to him, “In a certain city a pagoda is fallen. You must go and repair it, then write a dharanl [charm] and place it there. The reading of this charm will lengthen your life now and later bring you to Paradise.

The disciples of Buddha then asked him wherein

the power of the dharanl charm lay. The Buddha said, “Whoever wishes to gain power from the dharanl must write seventy-seven copies and place them in a pagoda. This pagoda must then be honored with sacrifice. But one can also make seventy-seven pagodas of clay to hold the dharanl and place one in each. This will save the life of him who thus makes and honors the pagodas, and his sins will be forgiven. Such is the method of the use of the dharanl. . . . The size of the pagodas shall be from an inch to a cubit in height or yet ten feet. From these pagodas, if the heart is set at rest by contemplation, shall come forth a wonderful perfume.” The Boddhisattva said, “. . . I will speak of the impressing of the law of the dharanl upon the heart. This dharanl is spoken by the ninety-nine thousand \oti15 of Bud¬ dhas and he who repeats it with all his heart shall have’his sins for¬ given. ... So shall ninety-nine copies be made of each of these dharanl, and they shall be placed within the pagodas. . . . These shall be honored with offerings and incense and flowers and there shall be a procession around them seven times while the dharanl is recited. Then will great salvation be wrought.”




In the face of the discrepancy in numbers between the directions given by Buddha and by the Boddhisattva, the empress evidently tried to be on the safe side and insure long life by ordering a million copies of the charm—and by so doing, she introduced printing to the world. The immediate purpose of her project failed, for she died about the time the pagodas were distributed, but the by-product of her act became one of the world’s greatest civilizing forces. It is typical of the international character which printing has always pos¬ sessed that this first printing project was in an Indian language in Chinese character and was carried out in Japan. In 782, thirteen years after the empress’ death, the great Emperor Kwammu moved the capital from Nara in order to escape the domi¬ nation of the Buddhist hierarchy, and the period of the domination of the state by the church was at an end. For two hundred years Japanese history is silent on the subject of printing—until the year 985,16 when it entered Japan once more as an importation from China. Meanwhile, printing in China itself had undergone a trans¬ formation.

NOTES 1. Chinese literature began to enter Japan as early as 540 (and may have been seeping in for a century before that), but Chinese civilization did not begin to make a strong impression until the beginning of the seventh century. Four Japanese students were sent to China in 608 to study, and on their return were instrumental in bringing about the great reforms of 645. 2. This was Yiian Chin-ch’ing who was given two titles in 776 and in 778. 3. Saeki, 1913:145. 4. Nihongi 23/8. For this and other information about Japanese seals, see Sporry, 1901:7-9. This passage from the Nihongi dates from about 720. 5. Commentary of Sho\u Nihongi \osho. 6. For a full description of these printed textiles, together with a bibliog¬ raphy of Japanese books on the subject, see Asakura, 1909:4. See also Toei shu\o, English introduction and plates.



[Ch. 7

7. See Ko\ushi dai-jiten (edition of 1916), p. 1395, and last plate in supplement. 8. Cf. Tsunoda and Goodrich, 1951:52, 64. 9. Sakanishi, 1939:147, in reviewing Kazuma Kawase’s studies on old movable type printing in Japan, wrote in part as follows, “After a painstaking comparative study, Mr. Kawase concludes that the characters of the dharani were written on a soft medium, probably a clay tablet, and a copper plate was cast from this tablet. Each casting brought about a slight variation in the impressions.” There is an earlier passage, dating from 75F which is claimed as a refer¬ ence to block printing in Japan, but the interpretation is uncertain. The pas¬ sage is as follows: “In the second year after the death of Otomo Akamaro [governor of the district of Tama in the province of Musashi, died 750] there was born a calf with black marks on its back. These marks had the appear¬ ance of an inscription on stone. They were interpreted to mean that Akamaro had appropriated to himself temple property and had died before punishment had overtaken him, and that as retribution he had been reborn in the form of this calf. At this all his family mourned deeply and feared, saying, ‘It is a fearful thing to commit sin. Can such a crime remain without punishment?’ This event was announced in a \atagi and in the sixth month of the same year was published abroad in order that those who should read it should repent of their sins and do good.” Nihon-\o{u gempo zen-a\u reii-\i, middle section. A \atagi in later writings means a block print. The fact that this event was only a few years before the first known block printing and that the statement here referred to was “published abroad” by this means, has led certain Japanese writers to regard this \atagi as a block print. The question is discussed by Asakura, who gives the text in full. Kawase discredits this passage; cf. also Sakanishi, 1939. 10. There is some confusion about the exact date of this event. The Empress Shotoku ruled for the second time from 765 to 769. The account in the Nihongi gives the year equivalent to 770 as the date when the printing o the charms was ordered. On the other hand the temple record Todaidi yoro\u gives the year 764 as the one in which the pagodas, containing the charms, were made and distributed. To add to the confusion, the name of t e ruler is here given as Koken, who reigned from 749 to 758. This same temple record gives 767 as the date when halls for the pagodas were built in the temple. For text of the Sho{u Nihongi statement and the temple record see Asakura, 1909:8. r , Sat°W’ l882’ Siv t^iat some °f the poetry of Po Chii-i (772-846) and of himself was printed. This preface bears a date equivalent to January 2, 825. Pelliot (i953:3I-33) vigorously disputes their understanding of Yuan’s words, while Dr. Hu (1954:122) as strongly supports it. Until archaeological evidence can be educed to settle the question, Yuan’s preface cannot be used as certain proof of printing in the early years of the ninth century. 16. Yiin ch’i yu i. The quotation appears in the three chiian edition of the Ssu pu ts’ung \’an hsii p’ien B/i8ab and in the twelve chiian edition of the Ts’ung shu chi ch’eng (No. 2833) 10/60. 17. Pelliot, I953:34_3^> points out that the original text, now written Ho-yii Ch’uan, has been transmitted faultily, as shown by the researches of Kuwabara. 18. Possibly


prince of the Liu Sung period,



19. Ssu-ma Kuang’s History, in the records of the T’ang dynasty, has the following account of Liu P’ien’s life: “In 902 Liu P’ien was made governor of Lu-chou. The Liu family from the time of Liu Kung-cho had been held in honor by scholars and officials for its adherence in every generation to filial, fraternal, and social duties. . . . The eunuchs hated him, and hence he was long punished by being kept in provincial posts. He wrote a book of admoni¬ tions for the junior members of his family.’ Translated by Meadows, in Curzon, 1860:16. 20. An astrological concept. 21. This statement is found in a book entitled Chia hsiin hsii by Liu P ien. It is quoted in full in the Older History of the Five Dynasties in an editorial note in the edition of 1739, 43/1—a note in which many of the earliest sources on the history of printing are gathered together. It is also quoted in full by Yeh Te-hui in Shu lin ch’ing hua I/18-19. In abbreviated form it is quoted by Yeh Meng-te in a book entitled Yen yu (Ts’ung shu chi ch’eng edition No. 2754-5) 8/74, written about 1130, which in turn is quoted by Ma Tuan-lin in the book Wen hsien t’ung l(ao. Other abbreviated versions are found in the Ko




chih ching yuan (39/23) and in Liu-an, 1916:2. European writers who have called attention to Liu P’ien’s statement are Meadows (loc. cit.) and Waley, I9I9:4I4- Cf. also the translation and comments of Pelliot, 1953:40-41. 22. This catalogue is reproduced in Dai-Nihon Bu{{yo zensho II, 99-106. 23. Kimiya Yasuhiko, 1932:27. 4 24- The ^ext may be found in his 7 ming chi (contained in the Piao sheng

wen chi, Ssu pu is ung {’an edition) 9/3, probably written in 887 or earlier. 25. Cf. Pelliot, 1923:269 and Hu, 1954:123. The Ching-ai ssu was erected in a.d. 582, according to

Honan fu chih

of 1867, 75/1 ra.

26. This is the plausible suggestion made by Robert des Rotours in a foot¬ note in Pelliot, 1953:44. Unfortunately for this theory there is no mention of burning in the official accounts. The Chiu Tang shu i9B/2ia bluntly records that the city was occupied peacefully.

Dr. Hu Shih

(letter to L.C.G. of

February 16, 1955) argues for an earlier date. Wu, 1936b: 141, considers instead that the reference is to the proscription of things Buddhist in 845. tunately Hui-ch’ueh is not otherwise known.


27. Other early authorities who mention printing in the Tang dynasty seem either to depend on the sources given above or to add little to our knowl¬ edge. The first is true of Chu I (1098-1167), in his 7 chiao liao tsa chi (edition of 1774) 2/61, who wrote: “Printed texts did not exist before the T’ang. At the end of the T’ang there were for the first time blocks for the printing of books at I-chou [Cheng-tu].

It is only under the Later T’ang [923-36] that the

Nine Classics were printed.

All [manuscript copies of the] classics and his¬

tones which were kept by private owners were assembled [to be suppressed], and it was held that the printed copies were the only correct ones.

See the

Liang ch’ao {uo shih [a large work of 120 chuan, now lost, by Wang Kuei el al, completed in 1082].” The same is true of the Kuo shih chih; frequently quoted in the thirteenth century Yii hai, which Pelliot, 1953:44-46, thinks may have been the bibliographical section of the above mentioned

Liang chao {uo shih. The second is true of Wang Tang (/?. 1101-25), a native of Ch’ang-an, who makes mention in his Tang yii lin (Ts’ung 'shu chi cheng edition, Nos. 2756-59) 7/203-4 of the “controversy over the accuracy of privately printed almanacs” when the Emperor Hsi-tsung arrived in Szechuan on February 3, 881. On these early references to printing see besides Pelliot, 1952, TN G

-oc 03

> ■. .




^ 3 c/> 2 •* -5 Pi

. P3 oj

^ u _n

WO .2ti Gr < u, 3

O, Cl, C

Ch. 9]



with pages but each page conforms closely to the style of the pasted sheets of the Tang manuscript rolls. Each half page contains eight columns, and each column either sixteen or twenty-one characters. The printed matter found in the Tun-huang caves and described in Chapter 8 comes in the main from just the time that Feng Tao was printing the Classics. Of the ten dated specimens at London and Paris, six (exclusive of duplicates) contain dates ranging from 947 to 950, and almost all the Tun-huang prints date from Feng Tao’s century. But this is not Feng Tao’s work. The work of the National Academy was Confucian—orthodox—whereas the Tunhuang finds are Buddhist. What the Tun-huang finds do reveal is that, side by side with the official Confucian printing of Feng Tao which so many literary men described, Buddhist printing continued to pursue its quiet course, ready to culminate in that great under¬ taking, the printing of the Tripita\a, which ushered in the Sung era, and which will be described in Chapter 10. An additional proof of this came unexpectedly in September, 1924, with the collapse of the well-known Thunder Peak Pagoda, or Lei Feng T’a, which for nearly a thousand years was a landmark of the West Lake region, at Hangchow. In the debris were discov¬ ered several tiny scrolls which had originally been inserted in holes bored in the bricks of the monument. These were printed sutras, entitled Pao ch’ieh yin t’o lo ni ching,24 and clearly dated Hsien-te i-hai, eighth month, or the early autumn of


975. The first lines

of the text report that the printing was done at the command of Ch’ien Shu (or Flung-shu, 929-88), ruling prince of the state of Wu and Yiieh during the years 948-78 and a staunch supporter of Buddhism.25 It is likewise mentioned that 84,000 rolls of the sutra were printed and deposited.26 This was exciting enough, but more was to come. Two or three years later two Chinese scholars, Wang Kuo-wei and Chuang Yen, announced that a short while previously there had come to light similar sutras of even earlier date, Hsien-te third year ping-ch’en, or a.d.

956. At least three copies of these precious scrolls are preserved

in the United States, one dating from 956 27 and two from 975.28 Another, dating from 975, is in the British Museum and others are



[Ch. 9

doubtless still treasured in China and Japan. Then in the spring of 1954 a significant discovery was made in Japan. When the Japanese Buddhist pilgrim Chonen (d. 1016) was in China during the years 984-85 he acquired a statue which he took with him to Japan. For nearly a millennium this has been one of the treasures of the Seiryo-ji, a temple in Kyoto. A few months ago it was opened for the first time by Professor Zenryu Tsukamoto and found to contain a cache of documents and other relics. The documents include a Diamond Sutra printed in 984 and three woodblock-printed paint¬ ings? which have been deposited in the Ueno National Museum, Tokyo.29 The work of Feng Tao and his associates for printing in China may be compared to the work of Gutenberg in Europe. There had been printing before Gutenberg—block printing certainly and very likely experimentation in typography also—but Gutenberg’s Bible heralded a new day in the civilization of Europe. In the same way there had been printing before Feng Tao, but it was an obscure art that had little effect on the culture of the country.

Feng Tao’s

Classics made printing a power that ushered in the renaissance of the Sung era. It is too much, however, to call Feng Tao the inventor. Not only had printing existed before his day—and printing which differed very little in technique from that which took place under his administration—but also Feng Tao had no part, so far as we know, in the technical work. He was the prime minister who saw the value of the new invention and gave the order to print on a large scale. His name has gone down in history as one of China’s great inventors, but his glory should be shared with others who did more than he to inaugurate the new invention.

NOTES The emperors of Shu during its first period of independence were Wang Chien (907-18), whose magnificent tomb was discovered in 1939 and opened under the direction of F£ng Han-yi; see Feng Han-yi, 1944:1-11; and Wang Yen (919-25).

During the second period, or Hou Shu, they were

Meng Chih-hsiang (934) and Meng Ch’ang (934-65),




2. Li Yii was a native of Shantung. In 925 the prince of Wei took him to Szechuan at the time of the campaign of the Later T’ang. Being a Hanlin scholar, he must have seen printed matter there. During the reign of Mingtsung (926-33) he became tsai-hsiang or minister. He died in 935, with the result that Feng Tao has received all the credit for the official development of printing.

3. The engraving of most of the Classics on stone in Later Shu took eight years, from 944 to 951, involving the I, Shu, Shih, three Li, Hsiao ching,

Lun yii, Erh ya, and part of the Ch’un ch’iu and Tso chuan. The Ku-liang and Kung-yang texts of the Ch’un ch’iu, the Meng-tzu, and two other works were completed in the eleventh and twelfth centuries. An astonishing feature of this work is that the Commentaries, as well as the official texts of the Clas¬ sics, were cut in stone, making an enormous work, greater in size than any other stone reproduction of the canon in previous centuries—more than 1,000 slabs in all. Cf. Kuei Fu, Li tai shih ching liieh 2/26-29, and Pelliot, 1953-55— 60. It ought to be mentioned that the brief text of the Lao tzu, or Tao te ching, was also frequently cut on stone in many parts of the empire during the T’ang. One such cutting was done in 736 at Ch’eng-tu. This and nine others like it engraved about this decade carry the commentary on the text written by










4. Possibly also the Western Tsin (265-317); cf. Pelliot, 1953:94. 5. Wu Chao-i was a native of Shansi province who first entered high official service in Shu in 935. Later he was detached to be superintendent of salt and iron—an important and probably lucrative post in Szechuan province. Around 944 he was back at the capital of Shu as minister, serving until 955. He died about 966-67. Cf. Pelliot, 1953:75-^1 • 6. Wang Ming-ch’ing (1127-95+), in the book Hui chu lu, quoted in the

Chiu wu tai shih 43/2^ editorial note. On Wang and the importance of his writings see Chang Chia-chu, 1940:199—210; also Wu, 1936^146. 7. On this famous collection of prose writings made by Hsiao T’ung (501-31) see Edwards, 1948:770-88. 8. As Wang Kuo-wei, 1923:144, points out, this is a blunder. The Wang family lost control of Shu in 925. Wu became a high official of Shu only in 935 when the Meng family was in power. 9. The National or Imperial Academy.

This text, while of interest, is

not entirely accurate, as we shall see, for Li £ s work as chief calligrapher in the first imperial printing of the canon began in 932, at least a decade before the Shu edition was undertaken. 10. Ssu-ma Kuang, Tzu chih t’ung chien 291, under date of Kuang-shun 3, fifth month, ting-hai, or June 22, 953. undertaken about 944 and concluded in

Wu Chao-i’s work was probably

953- Cf*

Pelliot, 1953 *79 ^1 *




11. The four dynasties were the Later T’ang, the Later Tsin, the Later Han, and the Later Chou. 12. Biographical notices of Feng may be found in both the Old and the New Histories of the Five Dynasties: Chiu wu tai shih i26/i-i2a and Hsin

wu tai shih 54/2a-5b.

Feng Tao, long condemned for serving so many

houses by orthodox Confucianists, was vindicated by the “iconoclast” Li Chih (1527-1602); cf. K. C. Hsiao, 1938:338-40. 13. The Classics were cut in stone during the T’ang dynasty at Ch’ang-an between 836 and 841, and are still in part preserved.

Photographs of them

have been published in Chavannes, 1909: Plates 356-64.

(On this work see

Gardner, 1938:61.) It is these Stone Classics that in general served as a model for the wooden plates made under Feng Tao’s direction.

The original con¬

tribution of the National Academy was to incorporate commentary with the text. The manuscripts on which the Stone Classics of 836 to 841 were based contained both text and commentary.

Of this only the text was copied on

stone. Feng Tao again included the commentary. 14. The only certain examples of printing in Wu (or Wu-Yiieh) date from some years later; they are the scrolls of 956 described below. 15. Ts’e fu yuan \uei 6o8/3ob-3ia. the original memorial.

This is obviously a condensation of

16. There are biographical notices of Ma Kao both in the Old and New Histories of the Five Dynasties: Chiu urn tai shih 71/53 and Hsin wu tai shih 55/8a. He flourished between 854 and 933. His ming is incorrectly written in

Sung shih 43i/26b and Yu hai 43/iob. 17. For the biography of T’ien Min (879-971), see Sung shih 431/263. 18. This account is abridged from the Ts’e fu yuan \uei 6o8/29b-3oa. This is the oldest and fullest account.

For those who wish to compare different

accounts of this event given by historians of the next three centuries, the fol¬ lowing translations are appended: Official history of the Later T’ang dynasty 43/23; “In the third year of the period Ch’ang-hsing (932), on the day hsin-wei (March 28), in the second month, the official known as chung-shu, or secretary of a department, made a memorial to the emperor, proposing to take as a model the characters of the stone inscriptions and to cut plates for the printing of the Nine Classics.” Official history of the Later Han dynasty ioi/6b: “In the first year of the period Ch’ien-yu (948), in the fifth month, on the day chi-yu (June 10), the National Academy sent a memorial to the emperor, stating that there were still four Classics


the Chou li, the 7 It, the Kung yang chuan and the Ku Hang

of which no plates had been prepared; and requesting that scholars be

called together to edit the text for the purpose of producing the plates. The petition was granted.” A °fficial !)istory of the L^er Chou dynasty 126/43, in the biography of Feng Tao: “In the time of Ming-tsung of the T’ang (i.e., Later T’ang)

Ch. 9]



dynasty, because the Classics had in them many mistakes, Li Yu, the officer in charge of education, with T’ien Min and others, taking as models the Classics as cut in stone by Cheng Tan (/?. 836-41) at the Western Capital, cut blocks for printing and thus spread the Classics abroad in the world. All who followed them had the work of these men as their foundation.” Official history of the Sung dynasty, section entitled Ju lin chuan 431/ 26b-27b: “In the beginning of the T’ien-ch’eng period (926-30), T ien Min, a doctor of the National Academy, directed Ma Kao and others to work with him in the revision of the Nine Classics. ... In the fourth year of the T ien-fu period (939), T’ien Min was given the title of chi chiu (libationer). . . . Though T’ien Min’s scholarship in the Classics was based on solid ground, he loved forced interpretations of passages. In the Nine Classics which he edited, he frequently put forward his own subjective interpretation.”

Ts’e ju yuan \uei 6o8/3ob-3ia: “Feng Tao, the prime minister of the Later T’ang dynasty, and Li Yii, wished to do honor to the ancient classical learning. They said, ‘During the Han dynasty Confucian scholars were hon¬ ored and the Classics were cut in stone in three different scripts.

In T ang

times also stone inscriptions containing the text of the Classics were made in the Imperial School. Our dynasty has too many other things to do and can¬ not undertake such a task as to have stone inscriptions erected. We have seen, however, men from Wu and Shu who sold books that were printed from blocks of wood. There were many different texts, but there were among them no orthodox Classics. If the Classics could be revised and thus cut in wood and published, it would be a very great boon to the study of literature. We therefore make a memorial to the throne to this effect.

The answer of the

emperor was that T’ien Min and other scholars were to examine and revise the text of the Classics and of the Commentaries. The work was carried on with zeal, and included the Book of Poetry and the three commentaries of the Ch’un ch’iu. The text was corrected, and the blocks were revised. Proofs were adduced with regard to the exact reading of the text, and the work was brought together in books. The Classics were first in this way made exact, and then they were cut in blocks. Money was appropriated from the Chengshih office [des Rotours, 1947-48:1, 11, calls this the “grande salle du gouvernement des affaires”], and also unappropriated money from various branches of the government was given out, as well as taxes from second degree graduates, in order to pay for the labor.” Ts’e ju yuan \uei 6o8/29b-3oa: “In the fourth month of the year 932, the following order was given by the emperor,

For the purpose of revising

the stone inscriptions of the Classics, uniting them with the Commentaries, and having them cut in plates for printing, it was recently ordered that from the National Academy specially qualified men be appointed, five or six for each one of the Classics, to examine the text and to add to it the Commentaries; now it is ordered that there be appointed from the court officials five men to supervise the work.

[Names of five officials including T’ien Min.]

Since the

establishment of the text of the Classics is of great importance, an importance not to be compared with that of all other books, although I have already



[Ch. 9

ordered the National Academy to appoint officers to edit the work, yet, because the work is so vast, and I still fear that errors may creep in, I order Ma Kao and the men with him (who are all great scholars and each one a specialist in the Classics), to make a final exact examination, in order that everything may be brought to absolute perfection.’ ” Parenthetically it may be explained that, during the last years of the T’ang, a half century earlier, there were on the staff of the Kuo-tzu chien five po-shih and three hundred students—sixty each for the study of the Five Classics (Chou li, I li, Li chi, Mao shih, and

Ch’un ch’iu with the Tso commentary). Cf. des Rotours, 1947-48:1, 447-48. Shen Kua (1030—93) in Meng ch 1 pi t’an iS/ya: “Under the T’ang dynasty block printing, though carried on, was not fully developed. From the time of Feng Ying-wang (Feng Tao), first the Five Classics, and then in general all the ancient canonical works were printed.” Ssu-ma Kuang (1019-86), Tzu chih t’ung chien 277/166: “In 932, in the second month on the day hsin-wei, orders were first issued to edit the Nine Sacred Books and print them for sale.” And at 291/93, under date of Kuanghsun third year, sixth month, jen-tzu (or July 17, 953): “Formerly in the time of Ming-tsung of the T’ang (Later T’ang) dynasty, the ministers Feng Tao and Li Yu prayed the emperor to command T’ien Min of the National College to correct the Nine Sacred Books, then to cut blocks for them and print them for sale, and the court assented. The present edition was printed and presented to His Majesty on the ting-szu day of the month (July 22, 953). From this time forth, even in periods of anarchy, the Nine Sacred Books were transcribed and diffused very widely.” Yeh Meng-te (1077-1148) in Shih lin yen yii 8/74: “Before the T’ang dy¬ nasty all books were manuscripts, the art of printing not being in existence. In the time of the Five Dynasties, Feng Tao first memorialized his sovereign, praying that an official printing establishment might be put in operation.” Chu Hsi (1130-1200) in T’ung chien \ang mu 56/326: “In 932 the T’ang (Later T’ang) dynasty for the first time cut blocks for the Nine Sacred Books and had them printed for sale.

(This quotation and the two preceding are from

the translation of Thomas T. Meadows in Curzon, 1860:1-33. There is some question about the rendering of the word here translated “for sale.”) Wang P’u (922-82) in Wu tai hui yao (Ts’ung shu chi ch’eng edition No. 0830, 8/96): “In the second month of 932 the chung-shu official wrote a memo¬ rial recommending that, with the Stone Classics as a basis, the Nine Classics be printed from plates. It was ordered by the emperor that the National Academy bring together leading Confucian scholars with their assistants, and that they should take copies [rubbings?] of the Stone Classics from the Western Capital; and that, each according to the particular Classic that was his speciality, they should copy and annotate the text, and then read them through with the minutest care; that then workmen of ability in the printing of characters be employed; that each department, following the model prepared, should cut the plates, and that the books thus printed should be spread abroad in the world.

If anyone should have a desire in the future to write a copy of the

Classics, it should be forbidden to do so except in accordance with these




printed copies; that it should not be allowed again to bring out miscellaneous editions. In the same year in the fourth month, the order was given by the emperor that the guest friend of the crown prince, Ma Kao, [and other officers, including T’ien Min and Ch’en Kuan, a po-shih or doctor according to the biography of T’ien Min in Sung shih 431] be appointed to have oversight of the work.” Wang Ying-lin (1223-96), Yu hai 43/iob-na: “In 932 in the second month, the emperor ordered the National Academy to revise the text of the Nine Clas¬ sics, and, using the books [rubbings?] of the stone inscriptions from the Western Capital, to have them copied and cut on wooden plates, in order to have them spread abroad through all the empire. In the fourth month, Ma Kao, Ch’en Kuan, and T’ien Min were ordered to examine the work with the greatest care. In the sixth month of 953, on the day ting-szu, the plates of the eleven Classics, together with the Erh ya, the Wu ching wen tzu, and the Chiu ching

tzu yang were finished, and T’ien Min presented them to the emperor.” This last sentence may also be found in the Wu tat hui yao 8/96, quoted above. The Erh ya is a dictionary, compiled about the third century b.c.; the second work, in three chiian, was written by Chang Ts’an (/?. 766-79) and carries his preface of 776; the last is a work of one chiian by T’ang Hsiian-tu (/?. 833— 40). 19. In 947 the Liao army entered Pien, or K’ai-feng, and removed, among other things, the classics cut in stone to the Supreme Capital. Liao shih 4/14315a; also Wittfogel and Feng, 1949:221, 256. 20. By piecing together the various accounts it would seem that the Nine

1 li, Chou li, Li chi, and the Ch’un ch’iu and its three commentaries. At the same time the following were printed: Hsiao ching, Lun yii, Erh ya, Wu ching wen tzu, and Chiu ching

Classics were: 7 ching, Shu ching, Shih ching,

tzu yang. The names of these two last, which were studies on the form of characters of the Five Classics and the Nine Classics respectively, may account for the fact that while most authorities speak of Feng Tao as having printed the Nine Classics, Shen Kua speaks of his work as the printing of the Five Classics. Both Five Classics and Nine Classics were conventional terms used at different times to refer to the Confucian canon.

(Compare our use of the words

“Pentateuch” and “Hexateuch.”) 21. (Carter gave this as the account in the Wu tai hui yao, but I cannot find it there. L.C.G.)

The Ts’e fu yuan \uei 608/306-313, gives additional

details, including an account of the charge of embezzlement that was brought against T’ien Min in connection with his management of this printing project, and how it was hushed up. According to the Yu hai 43/1 ia and the Ts’e fu yuan \uei 6o8/3iab, two years after the publication of the Classics (in the second month of 955) ’ an" It was the Ching tien shih wen in thirty chiian by Lu Yuan-lang (556^-627?). Chang Chao, min¬ other book was entrusted to T’ien Min for printing.

ister of war, was associated in the work. We read in the Yii hai 37/33a^ that



[Ch. 9

four years later (959) Kuo Chung-shu (918?—977) was charged with prepar¬ ing for the printer the Shang shu shih win in two chiian, which is a section of the Ching tien shih win. Cf. Pelliot, i9i6:II, 167. It was revised in 972 and again in 999 and printed both times. A large private publication which appeared about this time was the col¬ lected poetic and prose writings of a prime minister of the Later Chou, Ho ing (898-955)- Dr. Hu says of him (Hu Shih, 1954:123) that he copied his work

in his own fine calligraphy to be carved on blocks for printing. ‘Several

hundred complete sets of his collected works, each totaling 100 chiian in vol¬ ume, were thus printed for presentation to friends as gifts.’*’ See also his official biography in the Chiu wu tai shih 127/53-73. 22'. Th^ Tz“ chih t>unS chien 29T/8b. The engraving of the Classics on stone in Shu in 944-51 may also have had some effect on Shu printing. It is to be noted that both Ssu-ma Kuang and Chu Hsi, writing in the Sung dynasty, mention the early wide diffusion of printing in Shu and state that the Classics were printed there at the same time that they were being printed in the imperial capital. 23. This book was reprinted in Tokyo in 1884, in the collection Ku i ts’unaunder the title Ymg fu Sung shu ta tzu pin Erh ya. It has been discussed y Pelliot, 1902:316-17, and by Wang Kuo-wei, 1923:143-45. Both Pelliot and Wang Kuo-wei come to the conclusion that this is a Sung reprint of the Li £ original, and that it is probably a very exact reproduction of the original, but with the taboos changed. 24. Composed by Pu-k’ung (or Amoghavajra), born in Ceylon 705. died at Chang-an in 774. Cf. Hobogirin, fascicule annexe No. 1022. The Sanskrit title is



25. Cf. Chavannes, 1916:208-23. 26. Cf. Chuang Yen, 1926:331-32. . Jj7A ^ in 3 Kcollec,tion in New York. The first lines may be trans¬


W The general,ss,mo of the empire and prince of the kingdom of Wu and Yueh, Ch len Hung-shu, has printed the fan ch’Uh yin sun-a in fn’1° Iv / PrTS the.m f°r saf'k«Ping in precious pagodas. Recorded ,n the th„d year of Hs,en-t«, png-ch’cn." I, was found in a pagoda of the

?s;:7izn;irat Huchow

(on "'hici'see


It is remarkable that just a year earlier, in 955, eighty-four thousand miniature stupas in bronze were ordered made by the same prince (cf Cha vannes, 1916:140-41), and several have been reported (cf. Maspero


63 and Figs. 26-28). The one pictured by Maspero was preserved in the KuohisJofv 3h thC T°r °f tHu m0°Untain famous in Chinese Buddhist n‘ Jo n tai shan- Jhe number 84,ooo has its antecedent in the third century before our era when the Emperor Asoka is said to have built 84 000

stupas in h,s dominion for relics of the Buddha. See H. A. Giles 1923-39 and

Ch. 9]



Beal’s translation of the travels of Hsiian-tsang (602-64), Book 2, pp. 14647, or Waley, 1952:50; also Soper, 1939-40. These passages occur in the pil¬ grim’s description of Kapilavastu and Magadha respectively. The little stupas, some five or six inches in height, were (according to Hsiian-tsang) used by the Indians as receptacles for a written abstract from a sutra. Wang Kuo-wei, 1940:21/216-226, gives a full report on this printed sutra. His description of it includes the following data: 2^ inches in width, inner printed portion 1.95 inches, 338 rows of characters, 8 or 9 characters to each row. Quite rightly he points to the fact that in Japan nearly two cen¬ turies earlier one million little wooden stupas had been made, each containing a dharanl of the Mu-\u ]o-\o sutra. 28. One is in the Library of Congress; see Hummel, 1937:169-70. Another is in the Harvard-Yenching Institute library, Cambridge, Mass.; a third is in the British Museum. Lionel Giles, 1925:513-15, reported that it is 2-^5 inches wide, 6 feet 9 inches long, and contains 268 rows of characters. 29. Personal communication, dated Tokyo, December 9, 1954, from Charles S. Terry.


Chao K’uang-yin, one of the generals who had been contending for

authority during the anarchic period of the Five Dynasties, suc¬ ceeded in the year 960 in placing himself on the throne and united the empire under his sway. This was the beginning of the Sung dynasty, a period which rivaled that of the T’ang in cultural achieve¬ ment. The T ang dynasty had been a time of rapidly extending frontiers and of contact with the lands of the West, a period of freshness and youth, an era of lyric poetry and religious faith. The Sung dynasty, shut out from the West by the steadily encroaching nomads, was a time of ripe maturity. Lyric poetry gave way to learned prose great compendiums of history, works on natural science and political economy, of a character and quality such as neither China nor the West, except for a short period in Greece, had ever dreamed of. Religious faith gave way to philosophic specula¬ tion, and the great systems of thought were produced that have dominated China to the first years of the present century. In art the lofty tradition of the earlier period was carried on and brought to fruition, so that the greatest and best Chinese paintings that are now extant come from the period of the Sung. In invention, what the T ang period conceived, the Sung era put to practical use. The magnetic needle, used in the main in earlier times either as a toy or for the location of graves, was applied to navigation. Gunpowder, already known and used for fireworks, was applied to war during the Sung dynasty. Porcelain was so developed as to become an article of export to Syria and Egypt. A similar development took place in printing. From an obscure art at the end of the T ang dynasty, it was already making rapid 82] HIGH TIDE OF CHINESE BLOCK PRINTING


strides forward during the half-century interregnum. But as Feng Tao’s Classics were published only seven years before the first Sung emperor ascended to the throne, it was not until that dynasty had become established that his work bore fruit. The printing of the Classics was one of the forces that restored Confucian literature and teaching to the place in national and popular regard that it had held before the advent of Buddhism, and a classical renaissance followed that can be compared only to the Renaissance that came in Europe after the rediscovery of its classical literature and that there, too, was aided by the invention of printing. This is the reason why Feng Tao’s work has been considered of such importance by Chi¬ nese historians. Another result of the publication of the Classics was an era of large-scale printing, both public and private, that charac¬ terized the whole of the Sung dynasty. In quality the block printing of the Sung epoch has never been surpassed. The fine workmanship of these artist-craftsmen—beauti¬ ful calligraphy perfectly reproduced in print—sets a standard for all time. The importance of calligraphy to the booklovers of the day is shown by the fact that in almost all Sung editions the name of the calligrapher who prepared the copy is recorded in the colophon along with those of the author and the printer. This also was a time of improvements in the technique of printing of which the most noteworthy was the invention of movable type. But that new devel¬ opment must be reserved for discussion in Chapter 22. The advent of the Sung dynasty caused little change in the print¬ ing administration that had been organized by Feng Tao. The National Academy was still in charge of the work. The first books published by the government were further commentaries on the Classics, literary compendiums, and classical dictionaries.1 In the order for the printing of one of these earliest works it is expressly stated that the arrangements as to paper, ink, and expense should be the same as in the case of the Nine Classics. The printing of this work was in charge of a man from Szechuan.' The next important work was a voluminous commentary on the Classics in a hundred eighty volumes.3 In 1014 the Nine Canons and their Elucidations4 were considered to have mistakes and deficiencies, so they were



[Ch. io

revised and reprinted. A few years later (1030) the Elucidations of the Odes and History 5 were printed once more. The conquest of Szechuan (Shu) in 965 brought the printing of that province and of the central empire together. Wu Chao-i, who had been for years the patron of printing in Shu, and who, more than any other, had seen the possibilities of the new art for making literature available to the people, was found by the conquerors, an old man, living in retirement and obscurity. He was brought to the imperial capital, his printing blocks were searched out and again used, and from that day the printing that had circulated in Shu became current throughout the empire.6 By the end of the tenth century the printing of the great dynastic histories had started. This was a monumental work in many hun¬ dred volumes and its execution required nearly seventy years. Like the printing of the Classics, it was entrusted to the National Acad¬ emy.7 From 1063 little is heard of the National Academy until after the conquest of North China by the Kin Tatars and the removal of the capital to Hangchow. This was a time of constant warfare and frequent invasion. In 1139, four years after the setting up of the capital at Hangchow, a new edition of the Nine Classics8 from the old plates was ordered by the emperor. Because some of the plates were lost, the work was still unfinished in 1157. An order was then issued for the preparation of new plates where needed, and the edi¬ tion was soon complete. From one authority it seems that a part at least of the dynastic histories was printed at the same time.9 Meanwhile it is clear that private printing was gaining ground and spreading through the empire. Although the only records of this private printing are the title pages of the books that have been preserved, yet from these a certain amount of incidental and dis¬ connected information can be gleaned, as for instance the fact that the poet Ch en Ch i was also a publisher. Two hereditary pub¬ lishing houses figure, century after century, in these title pages, the Yin 11 family at Hangchow and the Yii12 family in Chien-an. This latter place was located in Fukien, very near the birthplace of the philosopher Chu Hsi. Here printers by the name of Yii were pub-] HIGH TIDE OF CHINESE BLOCK PRINTING


lishing books during the Sung, and continued down into Ming times, over four hundred years. In the year 1275, just a few years before the conquest of all China by Kublai, orders were issued for the collection of wood blocks in every district south of the Yangtze River.13 If this edict stopped printing, it could not have remained long in force, for books with the mark of the Chien-an printing house continued to appear for still another century.14 The question naturally arises what kind of books were printed in these private establishments.

Classics, commentaries, and his¬

tories seem to have been the favorite subjects, just as they were in the government printing office. The feeling of sacredness that in China has always surrounded the written or printed page—the feel¬ ing that impels men as a pious act to gather and burn printed scraps of paper and thus save them from being defiled—prevented the printing of any books not considered of great worth and dignity. There is evidence, however, that the practice began early in the Sung dynasty of printing the winning essays in the great national examinations.15 Local histories—histories of provinces and of cities— were also printed, probably in great number, judging from the number of those still extant.16 Likewise, and especially later in the dynasty, the field of printed literature grew constantly wider, includ¬ ing works on medicine, botany, agriculture, collections of poetry, and belles-lettres.17 Sung editions are rare and consequently valuable, but there are a few in the possession of each of the great libraries of the United States18 and Europe as well as in the Far East. A monumental work in 2100 volumes, published in China shortly before the first edition of this book appeared, consists of photographic reproductions of rare old books from Chinese private libraries. A study of the three hun¬ dred and more volumes in this collection that are photographed from originals of the Sung and Mongol dynasties gives a good cross section view of the kinds of literature that were popular with Chi¬ nese publishers during the time when William the Conqueror was invading England and the barons were wresting the Magna Charta from the unwilling King John, and while the Crusading princes were fighting with Saladin for the possession of the Holy Sepulcher.



[Ch. io

Histories are in the lead. Next come the collected works of various noted essayists, commentators, poets, and philosophers—large col¬ lections, running to many volumes each. A number of contempo¬ rary works on agriculture also appear.19 As for the commercial side of this early publishing, a certain amount of light is thrown on the subject by an extract from an old account book, dated 1176: For book of 1300 pages, including 30 pages of extra heavy paper at the beginning and the same at the end: Paid to printer for paper, paste, and labor. Per copy

1500 cash.

Rent of plates.1200 cash. Selling price of books. Per copy.8000 cash.20

In spite of its high culture, the Sung empire could not success¬ fully compete in warfare with the steadily encroaching nomads of the north. First the Khitans or Liao swept over Manchuria and the northern edge of the empire and held their domain for more than two centuries. Then the Kin, ancestors of the Manchus, overthrew the Liao and advanced still further, occupying for a whole century all that part of China which lies north of the Yangtze, and forcing the Chinese dynasty of the South, now known as Southern Sung, to move its capital to Hangchow. Finally in 1235 the empire of the Kin, and half a century later the Sung domain itself, went down before the conquering Mongols. But “China is a sea that salts every river flowing into it.” Each nomad people during its century or more of rule became thoroughly Chinese in culture and in ways of living—so much so that, first the Liao against the Kin, and later the Kin against the Mongols, stood as bulwarks of Chinese civilization attempting to hold back the new inundations of barbarians from the desert.21 As for the printing of that part of North China that came under the Liao domain, very little is known. A statement in the Liao annals to the effect that in 1056 “a college was founded and all the Classics and their commentaries spread abroad,” makes it possible that there was a Liao edition of the Confucian Classics.

As to

Buddhist literature we are on firmer ground. The Tripita\a, to-] HIGH TIDE OF CHINESE BLOCK PRINTING


gether with supplements, was printed between 1031 and 1062 (pos¬ sibly 1068) in Khitan script.22 Under the Jurchen, who established the Kin empire, there was more interest in printing. Dr. K. T. Wu indicates that the official, or National, Academy published more than thirty works in Chinese and fifteen in Jurchen,"3 and private concerns printed at least eleven. The center of this activity was at P’ing-yang, in the northwestern province of Shansi, where the government printing office was estab¬ lished in 1130, only five years after driving the Khitan from North China. One of the major works of the dynasty was a Chinese edi¬ tion of the Tripitaka, the blocks for which were cut over a twentyfive year period, in 1148-73. Of the complete set of nearly seven thousand rolls, 4,957 are extant today. Stored for centuries in a monastery in Chao-ch’eng, Shansi, they were rediscovered by the learned world about 1931 and were brought to Peking by the Com¬ munists with something of a flourish on April 30, 1949.24 Besides other printed books noticed by Dr. Wu it may be mentioned here that the Kin dynasty, like the Sung before it, printed paper money. This subject will be treated in the following chapter. The coming of the Mongols, who finished the conquest of North China in 1234 and of South China in 1279, wrought little change within China. The policy of the conquerors was to accept the cus¬ toms of the land as they found them. The printing of China under the Mongols was thus similar to that of the preceding dynasty, but the broadening of the scope of printed literature, already noticeable, became even more marked. Among the lines of printed literature recorded are a medical book,25 a large quantity of almanacs,20 and even a play.27 This last is of importance because fiction and the drama first became prominent in Chinese literature in Mongol times. They were both considered as very popular, almost vulgar, literature—if the word literature could be applied to them at all.28 The printing of a play marks a long step toward the popularization of printing. Another bit of popular printing that has been preserved is a woodcut entitled, “Beauties who from dynasty to dynasty have overturned empires.”29



[Ch. 10

From many sources we learn that quantity production became far greater during Mongol times than in the earlier period and that it became still greater during the century after the Mongols were overthrown—the century before Gutenberg. It was apparently dur¬ ing the early years of the fifteenth century, just about the time of Gutenberg’s birth and of the large-scale use of metal type in Korea, that Chinese block printing reached its highest point, so far as the annual number of books produced is concerned.30 But Chinese con¬ noisseurs find in the printing of this time a corresponding deteriora¬ tion in quality and technique. There are many references to official printing in the annals of the Mongols and of the early Ming emperors. When Kublai cap¬ tured Hangchow, the southern capital, he carried away with him to Cambaluc (Peking) all the blocks that had been used by the official printing office, as well as a large number of blocks from the province of Kiangsi.31 A government printing office having already in 12363" been opened in Cambaluc, the addition of all this material gave great impetus to its work. In 1293 this office was united with the Hanlin College and charged with printing certain books in the Mongol tongue as well as in Chinese.33 In 1329, a special office was opened for the translating and printing of Confucian literature in Mongol, and another office for printing the “sacred teachings of the imperial ancestors.” 34 Though many Chinese books of Mongol times are still extant, resembling in every way the books of the Sung era, very little in the Mongol tongue has been preserved. The Mongols were a totally unlettered people when they began their conquests. It was only after Jenghis’ career was well under way that he employed some Uigur scholars to reduce his language to writing. The Mongols, not being themselves a literary people, were content to patronize Chi¬ nese literature in the East and Arabic literature in the West. Inas¬ much as the translation of books into Mongol was a matter of national pride rather than of utility, editions were small, and little of Mongol printing has survived. One survival that is of special interest is a Mongol poem, found by Pelliot at Tun-huang. This is apparently an original Mongol work and not translated from the






$ (I $

I £^W&W

>! ^ *f f $4M£®L

ufai&atm< ^^nwMsgr^4^*

Hsiian-tsang. Printed in Japan in 1157 in folded book form, the form usually used for Buddhist scriptures, both in China and

A Part of the Diamond Sutra, as translated from Sanskrit to Chinese by


One of the Earliest Printed Books of Japan.

scw «o



Chinese. The four fragments of Mongol sutras found in Turkestan and the Buddhist prints and paper money found in Mongolia itself will be described in Chapters 14 and 16, on Turfan and on the Mongol Empire.

So far, our description of the printing of the Sung era has taken account largely of Confucian printing, the printing of the Classics, and of those historical and literary works that followed in the wake of Feng Tao's great publication. In this we have followed the Chi¬ nese records. But the printing of the Buddhists, the Taoists, and two minor religious groups—the Manicheans and Moslems—went on side by side with the Confucian and the secular. Of these the Buddhist deserves priority, for it was Buddhist printing that spread to Japan and to Central Asia; and in China itself it was Buddhists who printed the greatest single work of which we have record. During the years 972-83 there was published in Ch eng-tu one of the most monumental works that history records. This was the whole Buddhist canon, usually called the Tripita\a, which con¬ tained both the sacred scriptures that had been translated from the Sanskrit and a smaller number that had been written independently in Chinese. This collection consisted of 5,048 volumes covering 130,000 pages. It therefore required the cutting of 130,000 blocks. This massive work, together with additions, was reprinted fre¬ quently during the Sung.35 The fact that the printing of the Tripita\a is ignored by the official historians, who describe in such detail the work of Feng Tao, speaks eloquently of the lack of regard in which Buddhism was held by the ruling and literary classes during the Confucian revival. It was in Korea and Japan that the printing of the Tripita\a had its great effect. In 989 the king of Korea asked for and obtained a copy from the emperor of China.36 About the year ion his successor gave orders that it be revised and reprinted in his own dominions. The work occupied sixteen years.37 A copy of the Tripita\a now in Tokyo,38 of which only two of the 6,467 volumes are lacking, was



fCh. io

printed in Korea in 1458 and brought to Japan between 1469 and i486, and is said to be a reprint made at that time by order of the king, from the thirteenth century wooden blocks that had been carefully preserved in the Hai-in monastery.39 Meanwhile, in 985, a printed copy of the Tripitaka was brought from China to Japan by the priest Chonen.40 In describing this event there is used for the first time in Japanese literature the word sun-lion, printed book. For two hundred years—since the famous million charms of 770—Japanese records had been silent on the subject of printing. Either the art died out and had to be reintroduced from the mainland, or else it was too obscure to find a place in the annals. After the Buddhist canon was brought from China in 985, nearly a century more elapsed before the first mention of any book being actually printed in Japan. A Lotus Sutra was printed before 1080 and at least ten other works before 1200, one of them being a tenvolume edition of the Vidyamatrasiddhi, published in 1088.41 An¬ other was the Diamond Sutra, published in 1157—the favorite book with printers of China, Japan, and Central Asia. A portion of this edition is now housed in the British Museum.

There are other

siltras extant dating from 1206 and 1223, and from that time on an ever-increasing number. One of the first original Japanese Buddhist works also appeared in these years—the Ojd Ydshu, or Essentials of Salvation, by the learned monk Genshin (942-1017). It was printed in 1217, if not earlier, and was often reprinted.42 The books printed in Japan during the time corresponding to the Sung and Mongol periods—in striking contrast to those of China

are all Buddhist. They are as a rule works of merit, printed

in payment of a vow to succor a parent, relative, or friend in the next world—or else to obtain special merit for the one who bears the expense of the printing. They contain inscriptions of which the following is typical: “Moronafu on mature reflection sees that the faults of the present life are more than can be numbered, and to expiate the sins of boundless ages of past time is an impossibility. He has therefore undertaken the printing of this true doctrine, in order thereby to eradicate his accumulated guilt.” 43] HIGH TIDE OF CHINESE BLOCK PRINTING


With the inauguration of the Ashikaga regime in Japan (1336), and the expulsion of the Mongols from China (1368), there began the second great influx of Chinese culture into the island empire. For four hundred years, during all the cultural triumphs of the Sung period, China and Japan had scarcely met.

Now all the

stored-up energy of four centuries began to enter Japan like a flood, with the result that Japan was able to carry forward the great tradi¬ tion of Sung culture during the long period of conservatism that set in on the continent during the Ming dynasty. This epoch is marked by the first printing of Chinese Classics in Japan. The earliest Confucian work found by Satow in his researches, a copy of the Analects, is dated 1364. Ten more such books bear dates between that and 1400. Up to this time Japan had printed nothing but Chinese trans¬ lations from the Sanskrit. Now for two centuries it printed orig¬ inal Chinese works. It was not till the end of the sixteenth century, so far as is known, that the first original Japanese work of a secular character appeared in print—a copy of the Nihongi (Historical Records), printed with movable type. Buddhist printing spread not only east but north and west. The great quantity of Buddhist fragments of this period found by the German expeditions in the region of Turf an, the Uigur center, is described in Chapter 14. A smaller number of printed sutras of a similar character and in the same languages was found by Pelliot at Tun-huang, not in the sealed manuscript chamber that contained the Diamond Sutra, but in one of the other caves.44 A third source of Buddhist printing for this period is found in the discoveries of the Russian expedition of Koslov in Mongolia45 The excavation of the buried city of Kharakhoto brought to light a considerable number of printed books, evidently Buddhist, in the language of the Tangut, a people of Tibetan origin, who occupied northwestern China and a part of Mongolia during the two cen¬ turies preceding the conquests of Jenghis Khan.46 These Tangut texts have not yet been deciphered. With them, however, were found a dozen or more printed books in Chinese which are of special interest because, unlike the books of Turfan, they almost all contain dates.



[Ch. io

The oldest of the Kharakhoto books is dated May 16, 1016. Of printed books that bear a clear date, it is therefore the fifth in age which has so far been brought to light.47 Like the Diamond Sutra of Tun-huang and the tiny scrolls from the T’ien-ning ssu pagoda at Huchow and the Lei Feng pagoda at Hangchow, this is in roll form, the sheets being pasted together end to end. And this is an abridged edition of that same Diamond Sutra that was found at Tun-huang, though in a different translation not hitherto known. Each sheet shows in the margin the name of the individual by whom the block was engraved, while a final note gives the name of the patron who supplied the funds. As the patron and at least two of the engravers have been identified as men whose names appear in local histories in the province of Shensi,48 this is evidently a book brought from that province, and a sample of the Buddhist printing that was going on in China proper near the old capital. In contrast to that part of China which was under the dominion of the Sung emperors, and where all Buddhist printing was of more or less humble origin, a number of the Kharakhoto sutras contain the statement that they were printed by order of the Tangut empress and at her expense. Among these is another Diamond Sutra, printed in 1189. One small but highly popular siitra translated into Tangut, and also printed in 1189, was executed by order of the emperor, who caused a hundred thousand copies in Tangut and Chinese to be distributed.49 Additional light on the Buddhist printing of the Tangut empire is shed by a note in the official records of the Yuan dynasty. On November 29, 1294, it is stated, “orders were given to the Hsiiancheng-yuan to stop cutting the blocks for the Tripita\a in [the lan¬

guage of] Ho-si [Tangut].”50 But these orders seem to have been ineffective. The printing did not actually cease until 1302 when about thirty sets, in some 3,620 chuan each, were completed.51 There must have been everywhere in eastern Asia a passion for printing this great work with its five to seven thousand volumes. The special significance of the Kharakhoto printing, like that of the Uigur of Turfan, lies in the fact that Kharakhoto was on the direct route to the West, and that these printed books, spanning the



period from 1016 to 1352, were found not far from the original center of the empire of Jenghis Khan in one of the first regions that he added to his growing domain.

Taoist books from the Sung dynasty are rare, and the progress of Taoist printing is more or less obscure. It has already been shown (Chapter 2) that the part played by the Taoists in the origin of printing, though obviously of great importance, is more difficult to trace than the Buddhist and Confucian contributions. Taoist ac¬ tivity in Tang times is also suggested by the work of the Ho-kan Chi, in Liu P’ien’s statement (Chapter 8), and in the beginnings of playing cards (Chapter 19). As for the Taoist philosophers, during the T’ang dynasty the works of Lao-tzu and Chuang-tzu had been given a place in the Confucian canon on account of the Taoist affiliations of the dynasty. Though this position had been lost before Feng Tao published the Classics, yet in one of the commentaries that he published—a com¬ mentary written in T’ang times—the works of Lao-tzu and Chuangtzu were included among the Classics

and, as we have noticed

above, other Taoist classics were also printed by official order be¬ tween 1003 and 1014. This is, so far as is known, the first printing of Taoist classical literature—works printed not by Taoists but as a part of the Confucian canon. The publication of two editions of the Confucian Classics in 953 and of the Buddhist Tripit a\a in 983 evidently spurred the Taoists to competition. The compiling of the full Taoist canon was com¬ pleted in 1019—466 cases in 4,565 chiian—and was forthwith pre¬ sented to the throne.53 Nothing from this first edition survives although several reprints, made before the end of the dynasty, are still extant. Two such Taoist books are among Koslov’s finds at Kharakhoto,54 and one, a Sung edition of the Tao te ching, appears in the Ssu pu ts’ung l(an collection of photographic reproductions. Curiously enough, two books of the Manichean scriptures, adopted and taken over by the Taoists as their own, were printed



[Ch. io

along with the other Taoist books, either at this time or in one of the later editions of the Taoist canon. Further, there is evidence in the writings of Lu Yu (1125-1209) that Manichean treatises had been printed before the year 1000 in the province of Fukien, and possibly even a century earlier.65 That these sacred writings of a religion that made its way across northern Africa during the latter days of the Roman empire, and that until this century has been known to us almost wholly from the works of St. Augustine, should have found their way in the course of centuries into China, there to appear in print some two or three hundred years before the art of printing reached Europe, is one of the anomalies of history. One reason why few Taoist books have survived is the deter¬ mined attempt made by the Buddhists, often with imperial backing, to destroy them. In 1258 the Great Khan Mangu deputed his brother Kublai to represent him at a debate which was held in Kublai s presence between Buddhist and Taoist representatives with regard to the authenticity of the tradition concerning Lao-tzu’s ac¬ tivities in Central Asia. The Taoists were defeated, and the result was an imperial order to the head of the Taoist religion that he should bring all Taoist books to the capital and burn them, and that all the blocks for printing such books should be burned at the same time. This order not being completely effective, Kublai, now emperor of all China, issued a second edict decreeing the further burning of Taoist works.66 Moslems have never in any part of the world been fond of print¬ ing. Though Moslems entered China in great numbers during the Mongol period and of course brought the Koran with them, there is no record that the Koran or any part of it was ever printed in China. There are, however, records of printing on less sacred sub¬ jects. In the year 1328 there was a great issue of printed almanacs, 3,03,185 of them, in three different sizes, with special details about lucky and unlucky days for marrying, starting on a journey, making a garment, or buying goods. It is recorded that these almanacs were of two kinds, and that one kind was prepared specially for the use of Moslems.

There are also books on doctrine and liturgy, the



Moslem calendar, history, and geography, the Arabic language, and even a biography of Mohammed published in 1775-85.58 Of all the religions that flourished in China under the Mongols, there is just one that, so far as our records go, never printed. This is Christianity. Neither the Nestorians, whose churches Marco Polo found at Chinkiang and Hangchow and in Central Asia, nor the Roman Catholics, whose work flourished in Peking and Fukien soon after Marco’s departure, seem to have availed themselves of the new invention—at least, no Christian printing is recorded. How¬ ever, it must be remembered that even Buddhist printing is almost never recorded except in Buddhist books. Only the finding of these books has revealed the great part which Buddhist printing played. The discovery at some time in the future of a printed New Testa¬ ment from this period is not an impossibility and may throw light on the problem of how printing was transmitted. This story of the development and spread of block printing in Eastern Asia during the time of its greatest advance—during the four centuries before the art first made its appearance in Europe— may well close with a summary written by Yeh Meng-te about the year 1130.59 It is to be noted that the author wrote before quantity production really began, before most of the Sung books now in our libraries were printed, and before the speeding up that took place in Mongol and early Ming times. Before the T’ang dynasty, all books were manuscripts, the art of printing not being in existence. People regarded the collecting of books as something honorable, and no one had them in large quantity. Those who collected them had great ability in collating and compar¬ ing, whence it frequently happened that people had fine copies, and students, as a consequence of the great labor of transcription, also acquired great ability and accuracy in reciting them. In the time of the Five Dynasties, Feng Tao first memorialized his sovereign, pray¬ ing that an official printing establishment might be put in operation. And again in those years of our reigning dynasty called Shun-hua (990-94) °fficers were commissioned to print the historical records and the annals of the first and second Han dynasties. From that time forth printed books became still more numerous; scholars and officers ceased to make the collection of books a chief object of attention, and,



[Ch. io

as students found it easy to obtain books, the practice of reciting was in consequence broken up.

It is idle to speculate why in the succession of later centuries China lost the progressive educational impulse the existence of which this conservative writer so laments. Suffice it to say that every record of the Sung dynasty that we possess enables us to see a country which in its progressive thinking shows the result of this new stimulus and reminds the reader strangely of the reawakened Europe of the cen¬ tury that followed Gutenberg.

NOTES 1. The T ai ping puang chi, an encyclopedia in 500 chiian, was printed in 981; see Yu hai 54/353; the Shuo wen chieh tzu, a dictionary completed on January 29, a.d. ioo, and presented to the throne in 121, appeared in print in 986. 2. Kou Chung-cheng (929—1002) of I-chou or Ch’eng-tu was a pupil of Wu Chao-i; see Sung shih 441/xob—nb. The one in charge of editing was Hsii Hsiian (916-91), a native of Kiangsu; cf. Sung chih 441/6 ff. and Pelliot, 1923:18-19. 3. Wu ching cheng i, printed by Chao An-jen (958-1018), according to Sung shih 287/ 14b. This work deserves extended comment. It was compiled in over xoo p’ien, or sections, by K’ung Ying-ta (574-648) and others at the command of the second emperor of the T’ang. (See biography of K’ung, Hsin Tang shu 198/9!}.) The proposal to print was made in the third month of 988 by K’ung Wei (928-91). The then emperor put the task in the hands of the Kuo-tzu chien, and Iv ung Wei and others undertook the work of editing, while Li Yiieh and five others proofread it. (See Yii hai 43/153; in the biography of K’ung Wei the title is given as Wu ching su i.) Wood blocks were ready by the tenth month of 990. On the day ting-ch’ou of the first month of 998, Liu K’o-ming proposed that the work be corrected. On the day \eng-hsii of the following month this was done, and in 999 it was ready. (Yii hai 43/i5b-i6a.) The Wu ching cheng i is> usually cited in its several parts; viz., the 7 in fourteen chiian, the Shu in twenty, the Shih in forty, the Li chi in seventy, and the Ch’un ch’iu in thirty-seven, a total of 181 chiian. (Cf. Chiu T’ang shu 46/83, 9b, 10a, 12a, and 16a.) The Hsin Tang shu figures, given in 57/4b-i2b, differ: 7 in sixteen chiian and Ch’un ch iu in thirty-six, the rest the same, making a total of 182. We cannot account for the difference between x8x and 182 chiian given in the section on bibliog-

Ch. io]



raphy in the Histories of the T’ang, and the figure of 180 mentioned in the Yii hai 43/153. Wang Kuo-wei, i94ob:i/5a, has still another total, namely 179 chiian. His individual figures are the same as those of the Hsiti T’ang shu, except that he gives to the 7 only thirteen chiian. 4. Chiu ching chi shih wen. 5. Shih shu shih wen. 6. The full story is told in Pelliot, 1953:61-81. 7. The dates of the first printed edition of the dynastic histories are as follows (according to the Yii hai 43/16 f?.): Shih chi, former Han, and Later Han Three Kingdoms, Tsin, and Old T’ang Han chih North and South Dynasties, and Sui Liang, Ch’en, and New T’ang

994-1004 1000-1002 1022 1024-1027 1061-1063

The first three, together with the Nan shih, were corrected in 1034 and re¬ printed. A new revision of the Han shu was made in 1069. Cf. Yii hai 43/ 16b17a. The modern Po-na edition of the last is that of 1034-38. The dynastic histories that have come down from the period of Southern Sung (1127-1279) and have been reproduced in the Po-na edition include the Shih chi (printing of 1195-1200), Hou Han shu (largely that of 1131-62), and San \uo chih (largely that of 1190-94). Several others, Sung, Nan Ch’i, Chou shu, are based on an edition printed in large type in Szechuan about the year 1144. Cf. the notes of L. C. Goodrich in Yang, 1950:32-35. The exactness with which the dynastic histories were printed is described, no doubt with some exaggeration, by the Persian historian, Rashid-eddin. (See Chapter 17.) 8. According to the Yii hai 43/i8b-i9a, Five Classics were involved, not Nine. But another authority, an older contemporary of Wang Ying-lin (author of the Yii hai)—namely, Li Hsin-ch’uan (1166-1243)—informs us that what was printed were the Six Classics without the Li Chi, and the Dynastic His¬ tories without the Han shu. (Chien yen i lai ch’ao yeh tsa chi, quoted by Wang Kuo-wei, i94ob:i/4b.) Later the “Eleven Classics” were printed and the blocks stored for use. Cf., remark of Yiin Chen, dated 1305, quoted by Wu, 1950:465-66. 9. This is indicated by the list of missing blocks, one of which is from the Han shu. 10. For a full account of books published by Ch’en Ch’i and his son, see Yeh Te-hui, Shu lin ch’ing hua 2/28-31. The National Central Library, Nanking, was (in 1947) the proud possessor of an original edition of Ch’en Ch’i’s col¬ lection of Sung poets in 60 volumes as well as two anthologies also in the original edition. The director of the library, Chiang Fu-tsung, 1947:i~3> Puts them in the first half of the thirteenth century.



[Ch. io

11. Books printed by the Yin family are listed by Wang Kuo-wei, 1940b: 1/20a. Their shop was located on the main thoroughfare in front of the T’ai miao. 12. For a full account of the work of the Yii family at Chien-an, see Yeh Te-hui, 1920:2/13-18. Compare also Wu, 1950:473 ff. According to a report to the throne in 1775 of Chung-yin (d. 1778), then Manchu governor general of Fukien and Chekiang, it is not certain whether the Yii family of Chien-an were publishing books as early as the Northern Sung. (Yeh Te-hui, op. cit.) Wu, 1943:236 and 1950:489, informs us that the family continued their activities in printing throughout the Yuan and Ming dynasties into the seventeenth century. 13. The orders were issued on Sept. 29, 1275. See Yuan shih 8/27a. 14. The importance of the province of Fukien as a publishing center and the sort of work done in that and other provinces is thus described by Yeh Meng-te in Shih lin yen yii 8/74: “At present, of all books printed throughout the empire, those of Hangchow are considered the best, those of Szechuan come next, and those of Fukien are worst. Of late years the printing blocks of the capital begin to stand but little after those of Hangchow, but the paper used is not so fine. In Szechuan and Fukien, soft wood is much used for cutting into printing blocks, the object of which is their easy completion, with a rapid sale for the books. Hence the workmanship is not good. Fukien edi¬ tions are spread all over the empire, and that is on account of the ease with which they are got ready.” Translation of Meadows in Curzon 1860:15, 16. 15. A curious story, preserved by Kao Wen-hu (chin shih 1160), in the book entitled Liao hua chou hsien lu, Ts’ung shu chi ch’eng edition, No. 2867, 3-4, tells of two literary graduates from Shu, who, in the years 100816, purchased the printed essays of the prize-winners in the examinations. 16. Chu Shih-chia indicates in the first chart of his Chung-\uo ti-fang chih tsung lu that 28 Sung editions of various local histories exist today, and in Shih hsiieh men pao 1938:43°) he lists two more. In Sung times 177 titles are reported, according to Sung shih 204/i7b-24a. See Chu Shih-chia, 1950:15. Histones is perhaps a misnomer for these books. They are rather summaries of the official archives, such as have been published from time to time by every Chinese city. 17. An example in the first category is an early herbal, K’ai-pao hsiang ting pen-tsao compiled by Li Fang (925-96), which was printed from blocks in 973 or 974, and is unhappily no longer extant. Dr. A. W. Hummel (in a letter to L.C.G. of July 16, 1952) says that this information is based on a statement by Ts’ao Ho in 7 hsiieh tu chih, published in 1851. Other printed herbals appeared in 1061, 1108, 1159, and 1204. Cf. Sarton and Hummel, I94I:439~42) and Wu, 1950:455. The National Library of Peking likewise possesses a work on medical prescriptions, entitled Wai t’ai pi yao fang, compiled by Wang Tao in 752, and printed in 1069. See Hummel, 1946:22.

Ch. io]



18. In 1952 Princeton University’s Gest Oriental Library possessed 697 vol¬ umes printed from wood blocks cut in the thirteenth century; the Library of Congress had nine Sung printed works; Harvard University, two; and the Uni¬ versity of California (at Berkeley), seven “presumed to be Sung.” In 1954 Llarvard reported an additional three Sung works. 19. Entitled Ssu pu ts’ung /{’an, and published 1920-22. Fifty-six of the 323 titles represented in this collection are reprinted from originals of the years 960 to 1368. Since 1922 two other parts have been issued. The first part is described in some detail in the Report of the Librarian of Congress for 1923 (pp. 174-78). Among the Sung and Yuan editions are the following: Ssu-ma Kuang’s great history of China. Ssu-ma Kuang’s collected works (printed in 1133). Adventures of the Buddhist monk Hsiian-tsang in India. Illustrated description of crop plants (printed in 1204). Collected poems of Tu Fu, the T’ang poet. Collected poems of Su Tung-p’o, the Sung poet. About fifty other collections, each containing the complete works of some one author. 20. From T’ien lu lin lang hsii pien chih yii 2/4a—6a. This passage is taken from an account oiTa 1 ts’ui yen, a work of twelve chuan by Tseng T’ung of the Sung (the Imperial Catalogue 3/6a mistakenly accredits the work to Fang Wen-i of the same period and reports that it was of ten chiian)) it concerns the I, or Canon of Changes, according to the commentaries of the Neo-Confucians, and carries a postface of 1177.


21. The generalization that “each nomad people during its century or more of rule became thoroughly Chinese in culture and in ways of living is sharply disputed by K. A. Wittfogel, writing of the Khitan. (See Wittfogel and Feng 1949:7.) The latter’s view is sustained by Maspero and Escarra, 1952:119 ff. 22. See Wu, 1950:447-51. One of the first, and a justly famous item, is a glossary, which bears a preface of 997 by Chih-kuang, and may have been printed then. In 1064 the private publication of literature was prohibited. (See Liao shih 22/33.) One reason why the Sung Chinese knew so little about Liao publications is, according to Shen Kua, that the Khitan imposed an embargo on books, and anyone found guilty of taking them across the southern border was subject to the death penalty. For further discussion of the Liao Tripita\a, see Wittfogel and Feng, 1949:292-93. 23. Wu, 1950:453-59. This printing was done by the Kin National Academy, established in 1173, and the Hung wen yuan, established in 1194. Another office, the / ching so (Bureau for the Translation of the Classics), also translated classical and philosophical works, but there is no assurance that these were printed. 24. See report of Derk Bodde, 1950:197; he gives the size of the set now at the National Library of Peking as 4,330 chuan.



[Ch. io

25. Medical work of Sun Ssu-miao (581-682), printed about 1300. Among the Koslov finds from Kharakhoto were chiian 13 and 14 of Sun’s Ch'ien chin fang. 26. See Howorth, i888:I, 274; and Yule, 1903:1, 447. See also Cordier, I920:73~7427. Liu Chih-yiian chuan, printed about 1300; in the Koslov collection from Kharakhoto. The last sections (6, 7, 8) only were found. 28. Opinions have changed during the last generation. Wu, 1950:496, writes regarding the drama: “Previously regarded as vulgar literature, it developed under the Mongols into a fine medium for creative writing.” For more on printing under the Yuan see Wu, 1950:459-523. 29. In the Koslov collection from Kharakhoto. 30. In 1391, copies of the Classics and the dynastic histories were distributed among all the schools of North China. For additional particulars about early Ming printing, see Wu, 1943:203 If. 31. Kiangsi remained an important printing center during Mongol times. A list of Kiangsi books of the period 1312—21 is given in Liu-an, 1916:20—21; see also Wu, 1950:479. 32. Known as pien hsiu so. Another office known as ching chi so was opened at P’ing-yang. 33* The central government office for the printing of books was known by»different names at different periods in the Mongol dynasty, corresponding probably with slightly different functions. In 1236, under Ogatai, it was pien hsiu so; in 1273 under Kublai pi shu chien; in 1290 hsing wen shu; in 1330 i wen chien. 34. The office for printing the “sacred teachings of the imperial ancestors,” opened in 1330, was the \uang ch’eng chii. 35. In the eleventh to thirteenth centuries there were nine additional print¬ ings of the Chinese Tripita\a, the last one, made in the years 1277-94, num¬ bering 7,182 rolls. Cf. Yeh Kung-ch’o, 1937; Demieville, 1924 and 1953; Kenneth Chen, 1951; Pelliot, 1953:88-92; and Goodrich, 1953-54. A large portion of the edition of 1232 is housed in the Gest Oriental Library of Prince¬ ton University. See the report of the curator, Dr. Hu Shih, June 30, 1951, and his description of the library, 1954. There seem to have been other special printings of single sutras. Ellery Sedg¬ wick, 1949:6-9, in a whimsical account of the purchase of a statue made in Japan, tells the story of a find therein of a printed copy of the Lotus Sutra (see Nanjio, A Catalogue of the Buddhist Tripita\a No. 134), printed about 1160, probably in Hangchow, which he presented to the Library of Congress. 36. The request was granted in 989 (Sung shih 487/53) and the work received in Korea two years later. The Sung record is confirmed in the biography of the Korean envoy. Cf. Nak Choon Paik, 1951:64-65. (Demie-]



ville, 1924:193-95, writes that in 1019 another copy was presented by China to Korea. Paik puts the event in 991.) 37. The plates for this edition of 5,000 rolls were destroyed in 1231 or 1232. (Another edition or part of one seems to have been printed in 1047-83, but its history is obscure.) In 1236 a new edition was begun; it was com¬ pleted in 1251. The plates this time numbered 162,516. The Tripita\a now consisted of 1,511 separate works in 6,805 volumes. Each block—made of the pa\ tal tree wood—was about 91/2 inches long, 2 feet 2 inches wide, 2 */2 inches thick, and weighed approximately 8.26 lbs. The corners are metal plated to prevent cracking, and each varnished to make it insect proof. Nak Choon Paik, 1951:69-70. In addition to the Tripita\a itself, other books were published in Korea during the eleventh century. Tokujo Oya, 1950:197-202, informs us that Giten (d. 1101) traveled to Pien and elsewhere in China collecting books, which finally numbered approximately 4,000 and more fascicules. He estab¬ lished in the Tai-hong-pong sa a printing office to publish newly compiled Buddhist works by people of Korea (seventeen authors with thirty-eight works), of Liao (twelve authors with twenty-nine works), and of Sung (nine authors). On completion sets were presented to the empires of Sung and Liao, and to Japan. Some of these works were reprinted in China and Japan. 38. It survived the earthquake and the bombing of World War II, and is now preserved in the Zoja-ji, a temple in Tokyo. 39. Cf. Demieville, 1924:199. 40. Chonen received it in 984; Sung shift 49i/8b. See Tsunoda and Good¬ rich, 1951:56. 41. Cf. Peake, 1939:60. 42. Sansom, 1943:238-39. 43. Satow, 1882:53. 44. Pelliot, 1908:525-27. 45. Pelliot, 1914:503 ff. For further details the reader is referred to this article. 46. Wu, 1950:451-53, has also treated printing under the Hsi-hsia, or Tangut, empire which ruled in the west, with its capital at Ning-hsia, from 1032 to 1227. (Their state became autonomous earlier, in 982.) Translation and printing of the Buddhist canon into Tangut was begun in the eleventh century, but was not completed until 1302. Fragments of this huge work, in 3,620 chiian, have been found in the last fifty years in Ning-hsia and elsewhere. So too have other printed items, including two small glossaries, published in 1132 and 1190 respectively. 47. At Tun-huang dated fragments and single sheets earlier than 1016 were found and at least two undated books that are generally considered to have been earlier, but the Diamond Sutra of 868 and 984 and the To lo ni ching of



[Ch. io

956 and of 975 are the only dated books that are older. Liu-an’s reference (1916: 6) to a Diamond Sutra of 961 from Tun-huang seems to be due to confusion with the Diamond Sutra of 868. 48. The patron was an official at Tan-chou in Shensi (now I-ch’iian). The two engravers were from two subprefectures which are now included in the prefecture of T’ung-chou in the same province. Pelliot, 1914:507-8. 49. This was a text which included a description of the paradise of Maitreya, originally rendered into Chinese in a.d. 455 by a cousin of the em¬ peror of Liang. Cf. Nanjio, 1883: No. 204. Ivanov translated the introduc¬ tion into Russian and Chavannes into French. Chavannes, 1911:441-46. 50. Yiian-shih 18/46. See Pelliot, 1914:518. 51. Attributed to Kuan-chu-pa. Cf. Daijo Tokiwa, 1939. 52. Ching tien shih wen; on this work see above, Chapter 9, note 21. We also read in the Yu hai 43/196 that Tu Hao (938-1013/4) and others were ordered in 1003, in the fourth month, to edit the Tao te ching. It was finished in the sixth month. In 1005 the Chuang-tzu. was undertaken. The Lieh-tzu. and Meng-tzu followed in 1011-14. 53. Pelliot, 1912:141-56, and 1931:150. 54. One is a commentary on Chuang-tzu, an edition apparently of the thirteenth century—a book lost and unknown in China from the fourteenth century on. The other is No. 372 in P. Wieger’s list, 1911. 55. Chavannes and Pelliot, 1913:300-2. 56. Chavannes, 1904:385-87, 395-403. 57. Howorth, 1876:1, 274; Yule, 1903:1, 448. 58. See the list compiled by Vissiere in D’Ollone, 1911:393-417; to which add Pelliot’s comment, 1935:279. 59. Yeh Meng-te, Shih lin yen yii 8/74. Translation of Meadows in Curzon, 1860:14-15. The Chinese text is quoted by Yeh Te-hui 1/24.


The form

of early printing that was most widespread in China—

the printing that touched all classes of the people, and also attracted the attention of Marco Polo and other European travelers—was paper money. Some Chinese writers, in treating of paper money, consider it to have been a natural development from other forms of representative currency. They frequently refer to “white deerskin money” which the Han emperor Wu in


b.c. compelled all nobles to buy and

then present to him as tribute; but, as Dr. Yang remarks,1 the deer¬ skin was not intended for circulation, and so cannot properly be considered in this context. A more likely beginning starts with the first years of the ninth century, when “flying money” or certificates were given to merchants depositing cash in government offices called chin-tsou yuan.2 These certificates guaranteed reimbursement in places designated by the merchant. Whether these certificates were made of paper or not is unknown.

The official sources are silent on the point, but the

assumption is not unlikely as both the Sui and T’ang histories record the use of paper for currency in the years just prior to 618.3 Paper money for use at funerals—probably an imitation of metallic cur¬ rency—is also recorded in the biography of an eighth-century figure.4 “Flying money,” later called “credit cash,” was in use for some cen¬ turies, and so far as is known was not printed. The first tentative steps taken in the direction of genuine paper money may be credited, like so much in the early history of print¬ ing, to the provincial capital known as I-chou (the Ch’eng-tu of today) in Szechuan. The record runs as follows:5 After the rebellion of Li Shun in 994 the offices in several districts of Szechuan, includ103




ing I-chou, stopped making iron money for a season. This brought about a shortage of currency among the people and helped the circulation of paper chiao-tzu, or “exchange media. But the shops which handled chiao-tzu had no financial backing, so the govern¬ ment decided on supervision. In 1005 the prefect of I-chou, Chang Yung (946-1015), was concerned with this supervision, but is not— in the opinion of Ichisada Miyazaki—the creator of the system of supervision. After ion, government control became really effective. The chiao-tzu were then declared good for three years. At the end of this period they had to be converted either to currency or to fresh chiao-tzu. This system made supervision easy. At this time sixteen rich families in I-chou were given the right of monopoly over the chiao-tzu. For this right they had to bear some obligation towards the government. But these families had their troubles. Forgery became common and loss of confidence ensued. Thereupon the government decided to do away with the privilege and once more circulate iron money. Nevertheless, in spite of some lack of confidence, the people con¬ sidered the chiao-tzu necessary.6 Iron money depreciated in value and silver money was considered too costly. So the government in 1023 brought an end to the private system and in the following year, after due preparation, took it over—to be precise, on the twen¬ tieth day of the second month, or April 1, 1024. Prior to 1024, the question of government control of the system was frequently argued. Around 1016 Hsieh T’ien, then a censor of the I-chou circuit,7 made the first proposal, but the court refused to act. When K’ou Chen 8 became prefect of I-chou he asked for per¬ mission to discontinue the private handling of chiao-tzu, and this was granted. He went further and asked the court for the power to suppress all chiao-tzu, but this was not done. After K’ou’s transfer, the use of chiao-tzu revived. These notes were printed with copper seals fashioned by govern¬ ment order. In one year there was issued a total of chiao-tzu amount¬ ing to 3,884,600 \uan.9 In his description of the earliest privately circulated notes in I-chou, Li Yu specifically states that the printing was done on paper of one kind or color.10

East Asiatic Library, Columbia University. Bronze Block for Printing Currency in the Time of Kublai Khan (c. 1287).

(20.3 cm. wide, 28.25 cm. long, 0.63 cm. thick, resting on 4 pegs 1.6 cm. in height.)


mo rr nr FF M 3



5^ FT re-FT* gg Si?

c/D rt> 03


*- £

o OO i -T3 X

rrznrgg^^ kvtoiK. fT rvTzr nr pit

•t# E EfS tv f&rr g ^55f£ ~^wsrowRjgfl

Mill raRrlB^ ®, gives the dimensions of the copper plate of 1214 as .192 m. by .105 m. We know that the Kin dynasty issued paper notes at several times and in several denominations (cf. Yang, 1952:58—61), but is it likely that any one of them was identical in size and shape with the large ones issued in the Hung-wu period? 24. Professor Basil M. Alexeiev of the Soviet University of Leningrad very kindly examined and described the Mongol notes found by the Koslov expedi¬ tion. Sogabe Shizuo, 1951, illustrates one of them. 25. Bergman, 1945:149. 26. See Goodrich, 19503:127-30; and Sogabe Shizuo, 1951: Plates 2 and 3. 27. Dr. Yang 19526:65, makes this interesting observation: “Together with paper notes, Chinese banking practices became known to the West. Max Weber states that the accounting system (Verrechnungswesen) of the old Hamburg Bank was set up on a Chinese model. Robert Eisler suggests that the old Swedish system of banking and money deposit vouchers may have been influenced by Chinese examples, transmitted by medieval merchanttravelers and, possibly, by Jewish silk merchants.”




the westward movement of printing it is necessary

to form some picture of the ways by which culture passed between the West and Eastern Asia. The idea that the history of China has until recent times been a closed compartment, affected by noth¬ ing and affecting nothing in the Western world, is being rapidly dispelled by investigation. Each new journey of exploration into Central Asia and each new study of ancient literature makes it pos¬ sible to follow a little further the silken thread that has bound the civilization of the West to that of the distant East. Imperial Rome wanted silk, China had it. Here is the key to the development of a great caravan route that crossed Turkestan, Persia, and Syria, and reached the Mediterranean at the ports of Phoenicia and Palestine. Modern scholarship has not yet answered the question why the birth of thought—in the age of Confucius, of Buddha, of the Hebrew prophets, and of the early Greek philosophers—came to the widely separated lands of ancient culture at the same time. Much less has it explained the earlier Neolithic precivilization that developed similar stone implements and even similar pottery designs at various places along the route from Greece to China.1 With the establishment of Roman dominion in the West and the Han dynasty in China, the connection between East and West first begins to emerge into the light of history. Somewhere about 170 b.c.

a tribe known to Chinese annals as the Yiieh Chih, and later

to the Greeks as Indo-Scythians, a people possibly of Indo-European tongue, living within the borders of China in what is now the prov¬ ince of Kansu, left their home and moved westward. Within a little more than two centuries they had conquered the eastern provinces



[Ch. 12

of what had been Alexander’s empire, and had shown their ability to absorb diverse elements of culture by striking coins in Greek style, bearing the effigies of the gods of Greece, of Persia, of Egypt, and of India, and even portraits of Augustus Caesar and of Buddha. All the gods, including Buddha (who looks strangely like Apollo), are clearly labeled in Greek. It was in this Indo-Scythian empire that Buddhism was transformed to suit its more cosmopolitan environ¬ ment and the new Buddhism started on its long journey eastward to China and Japan. At the court of one of the Indo-Scythian kings, some time before their Buddhist and coin-striking days (ca. 128


Chang Ch’ien, an emissary of the Chinese emperor, gained for China its first clear reports of the lands of the West. Chang Ch’ien also brought back the seeds of alfalfa and the grape vine, which were planted in China by the Emperor, and are, so far as is known, the first plants introduced into China from the West.2 In the wake of Chang Ch’ien’s mission came the Chinese conquest of Eastern Tur¬ kestan, opening up the pathway across the Indo-Scythian kingdom to the Roman Orient, and with this conquest came an enlarged silk trade. Armies, ambassadors, and caravans were sent frequently to the West. One Chinese embassy in the year


97 reached the

Persian Gulf and was deterred from going on to Rome only by the reports they heard of a “kind of homesickness which men have when they are long upon the sea.” The first travelers recorded from Rome to China came by sea as far as Tonking in the year



and were led from there overland to the Chinese capital, Loyang. They are known in the Chinese annals as envoys from the Emperor An Tun, who has been identified as Marcus Aurelius Antoninus.3 During this period silk was the chief article of export from China, the Chinese keeping the process of silk production a strict secret. The first Roman writers thought of silk as a vegetable prod¬ uct, which was stated by Virgil to be combed from trees. Only in the second century of our era did they become better informed; Pausanias (ft. 117-180) described the silkworm with fair correct¬ ness. Silk came into the Roman Empire in ever-increasing quan¬ tities during the classic period, and continued to come into Con¬ stantinople after Rome had fallen. The re-opening of the silk routes

Museum fiir Volkerkunde. Leaf from a Sanskrit Book Printed to Imitate the Indian Pothi.

Sanskrit manuscripts

and even printed books were often made up like the ancient books of India that had been written on palm leaves. Such books are known as pothi. When the folded book came into use and the hole in the center was no longer needed for binding, a hole, often highly con¬ ventionalized, was still printed.

About twelfth or thirteenth century.

(15.9 X 31 cm.)

Museum fiir Volkerkunde. Bit of Tangut Printing.

The Tanguts were a people akin to the Tibetans who set up a

strong kingdom on China’s western frontier during the eleventh and twelfth centuries. Their kingdom was destroyed by Jenghis Khan soon after 1200. the Tangut language have been found at Turfan, (12 X 13.6 cm.)

Block printed siitras in

at Tun-huang,

and at Kharakhoto.

Museum fur Volkerkunde.

Fragment of a Printed Sutra in the Mongol Language in Square (’Phags-pa)

The page numbers are Chinese. Turfan.


Found in the ruins of the old city wall of Chotscho near Dates from about 1300. (14.2 X 20 cm.)

Ch. 12]



was one of the central features of the foreign policy of Justinian (a.d.

527-65) and his immediate successors. These routes had been

closed by the Sassanian power in Persia out of fear of the fast¬ growing Turkish kingdom on the northeast, first heard of in Western history at this time. Justinian tried to interest the king of Abyssinia and certain Christian princes in India in a project to open a new trade route with the East that should avoid the Persian dominion altogether. When this plan fell through owing to the lethargy of the king of Abyssinia, advantage was taken of an em¬ bassy from the Khan of the Turks, and a return embassy was sent in 568 by a route north of the Caspian Sea to the Turkish court (in Turkestan) and an alliance formed, the purpose of which was to compel Persia to allow a resumption of the silk trade. Meanwhile some Nestorian priests returning from the East had brought to Justinian the news that silk was produced by caterpillars, whose eggs they believed they could obtain. With the emperor’s encouragement they proceeded either to China, or—more likely— to the kingdom of Khotan in Chinese Turkestan, whither silk cul¬ ture had been brought in the year 419 by a Chinese princess.4 To avoid detection as they carried the precious eggs over the frontier, they hid them in the long bamboo staff that one of the priests carried. From the eggs carried in this bamboo staff—if the story told by the Greek chroniclers is to be credited—are descended the silkworms that have been reared in Europe down to modern times.5 The century after Justinian saw marked changes in the face of Asia. Two great empires, that of the T’ang dynasty in China and that of the Caliphs in the West, had divided most of Asia between them, and the two empires had met in Turkestan. From this time on it was the Arabs who supplied Europe with silk. Imported silk they drew from China through Samarkand. Silk culture they learned from Constantinople. Throughout the Middle Ages Europe bought the bulk of its silk—both Chinese silk and silk of Arabic manufac¬ ture—from the Arabs. Not until the end of the first millennium did the art of silk culture become known in Western Europe, being first introduced into Italy in the tenth century.



[Ch. 12

Silk was thus, so far as is known, the first of China’s great gifts to the West, reaching Europe some time before the Christian Era; but the art of producing silk, unlike most Eastern arts, reached Western Europe before the Crusades.6 Today there are few parts of the world less known and less often traversed by civilized man than the lands that lie between China and the Near East. This is the one great trade route of the world where the means of communication are poorer today than they were one or two thousand years ago. Other ways of commerce by sea and land have been opened and have left the ancient route from China to the Western world scarcely even a memory.7 It is hard today to imagine that great highway over which the world’s long distance caravans plied to and fro. Two or three years were spent on the journey—often more, for it is seldom that one man or one caravan made the whole trip. Men of all races and creeds relayed the silks of China and the varied wares of the West stage by stage over the long road. Great cities grew up, both in the fruitful lands and in the oases of the desert along this now deserted trail. Turfan and Capernaum, cities at the two extremes, that once profited by being near the great trade route, today are ruins. Samarkand and Baghdad have lost their glory. But there was a day when that whole road lay through the lands of prosperous peoples who gathered together the elements of culture from all the East and all the West, an eclectic and cosmopolitan culture that has been buried and preserved wherever the route lay across what is now desert, espe¬ cially in Chinese Turkestan. Nor was it through a short period that this prosperity of the cities along the Silk Ways continued. Whether ruled by Caesar or Caliph, Western Asia loved to clothe itself in silk, and it had prod¬ ucts that it was willing to send in exchange. Great empires—the Roman, the Indo-Scythian, the Caliphate, and the T’ang of Chinafacilitated the trade, which continued to grow until finally, just before the Crusades, Europe broke through and began to have her part also in the traffic along the Great Silk Ways. Owing to the fact that until the later Arab period goods were generally carried by relays and few caravans went the whole dis-

Ch. 12]



tance, people at one end of the line knew very little about people at the other. But that did not prevent ideas and products from making the long journey, even though the recipients seldom knew from whence they came. Pliny called the apricot and the peach “Armenian tree” and “Persian tree,” little knowing that the Arme¬ nians and Persians had received them from China.8 Aristophanes called chickens “Persian birds” without realizing that chickens came from Burma.9 When Saladin made a present of porcelain to the Sultan of Damascus, he called it Chinese, but when some centuries later porcelain began to be manufactured in Venice, it was called Arabic. So during the long period from Roman times down through the Middle Ages there was a steady give and take. The peach and the apricot, silk and tea,10 porcelain and paper, playing cards, and prob¬ ably gunpowder and the compass were among China’s gifts to the West. The grape and alfalfa, the carrot,11 glass manufacture,12 Nestorian Christianity and Mohammedanism, the alphabet,13 and some impulses of Greek art were a few of the things that Eastern Asia re¬ ceived in return.14 Laufer traces the history of some twenty-four agricultural products, the knowledge of which was carried westward from China to Persia or beyond from the Christian Era down to Mongol times, and sixty-eight that were carried in the opposite direction. The greatest gifts of Southern and Western Asia to the Far East were their religions, and these formed the closest cultural link be¬ tween East and West.

The advance of Buddhism from India

through Central Asia to China and Japan is well known. Less known but of great importance as forming a point of connection across the continent is the advance of Nestorian Christianity and Manicheism. The latter, the religion of Mani, which was founded in Persia during the third century on a substructure of Zoroastrianism and Gnostic Christianity, and which greatly influenced the thought of the Roman Empire (St. Augustine himself being a Manichean be¬ fore he became a Christian), has been little known until recent discoveries in Chinese Turkestan have brought to light a large quan-





tity of Manichean scriptures and other writings in Persian, Sogdian, Chinese, and certain dialects of Turkish. It is now known to have been the state religion of the Uigurs, whose capital was at Turfan. Certain of the Manichean scriptures were printed in China before the year 1000 (see Chapter 10). Christian missionaries of the Nestorian sect came from Persia and Syria into Chinese Turkestan sometime in the fifth or sixth century. In almost every site excavated in Turkestan, remains of Christian churches were found, with manuscripts in Syriac and Persian as well as in Chinese and the languages of Central Asia. Even the correspondence of some of these priests with their mother churches in Syria has been unearthed. Discoveries tend fully to con¬ firm the record contained in the famous Sian inscription of the introduction of Christianity into China in the seventh century and of its persistence, both in Central Asia and in China, down to Mongol times. The metropolitans of Central Asia and of China were subject to the Nestorian patriarch of Baghdad, a dignitary who, strange to say, was given by the Moslem Caliphs great freedom in the exercise of his authority. By special dispensation during the latter part of the period, the metropolitan of China was relieved of reporting to his superiors in Baghdad except once in four years. At one time (during the Mongol period) a Christian of Turkish extraction from northern China was made patriarch of Baghdad.15 The exaggerated reports that reached the Crusaders of the exploits of Prester John have now been traced as referring to the Nestorian king of one of the tribes of Central Asia. The Moslem penetration of the Far East began within a few years after the death of the prophet, when about the year 652 the first Moslem envoys reached the Chinese court. From that time Arab trade with China steadily increased. How the early Arabic trading posts flourished is indicated by Abu Zeyd, who, writing about 916, stated that in the rebellion of 878 in the city of Canton,16 one hundred twenty thousand Moslems, Jews, Christians, and Mazdeans, who were there on business of traffic, were killed. Even al¬ lowing for Arab exaggeration, there is evidence here that trade be-

Ch. i2]



tween China and the Saracen Empire had already reached large proportions. The infiltration of religious ideas from the West is again illus¬ trated in the account of the Arab traveler Ibn Wahab,17 who visited China in the latter part of the ninth century and described his audi¬ ence with the emperor. The emperor, after discoursing with con¬ siderable accuracy on the five great kingdoms of the world—the Chinese, Turkish, Indian, Arab, and Greek—is said by the Arab narrator to have pulled from a box beside his throne pictures of Noah in the Ark, of Moses and his rod, of Jesus upon an ass, and of the Twelve Apostles. The surprising modernness of this Chinese emperor as seen by his Moslem visitor is illustrated by the fact that though he marveled at what Jesus accomplished in the short space of thirty months, he combated the idea that there had ever been a universal deluge and laughed heartily when his Arab visitor tried to tell him that the world had been created only six thousand years. The Moslems in China always retained a close connection with their home base either by sea or across Central Asia. They were under a system somewhat similar to extraterritoriality, and it was not until after Mongol times that they began to be submerged as an integral part of the Chinese people. The large number of Moslems in China today, who as a rule are of mixed Arabic ancestry, shows the extent of this early infiltration and indicates how close must have been the contact between China and the West that was thus established. In return for religious ideas, which were moving eastward and northward across Asia, China sent back her inventions. Some of the inventions with which the Chinese have been credited still await the research student and nothing clear and definite can be stated until much careful study has been made. The invention of paper, which has been more fully studied than the others, is described in Chapters i and 13. Though gunpowder and the compass are still obscure in their beginnings, a few words about them as well as about porcelain may present some useful analogies to the student of printing.



[Ch. 12

It is known that explosive powder was used in the T’ang dynasty, though probably not for warfare. Its first use in battle was in the form of explosive hand grenades, or grenades thrown by various mechanical devices. When the use of these grenades first began is not certain. They are apparently described in a military handbook published in 1044 and may have been used in the year 1000 against the Khitan, and again in the wars against the invading Jurchen in 1126-27 and 1161-62. Following the Mongol conquest of much of Asia the Arabs became acquainted with saltpeter sometime before the end of the thirteenth century. They called it Chinese snow, as they called the rocket the Chinese arrow. Roger Bacon (ca. 1214 to ca. 1294) is the first European writer to mention gunpowder, though whether he learned of it through his study of Arab lore or through his acquaintance with de Rubruquis, the Central Asiatic traveler, is uncertain. All that can be said with assurance is that the use of gunpowder in warfare became known among the Saracens and in Europe very quickly after its first use in warfare in China.18 With respect to the compass, the Chinese had known the prop¬ erties of loadstone since before the Christian era, and during the first millennium after the Christian era there are many curious stories, the interpretation of which is still unsure (but which seems to point to the development of the differential gear), with regard to the construction of “south-pointing chariots.” 19 The earliest clear mention in Chinese literature (or any literature) of a magnetic needle is by Shen Kua (1030-93), the same man who first described movable type printing. The first mention in Chinese literature of the use of the compass for navigation is a little after 1100 but refers to the period from 1090 to 1102. At this time, according to the statement of Chu Yii, it was used by Chinese navigators between Canton and Sumatra. The first men¬ tion of the compass in Europe is in a poem by Cuyot de Salins about 1190 and again around 1204 by Cardinal de Vitry, who visited Palestine in the fourth Crusade, and who describes loadstone as having been brought from “India.” The indications would seem to be that the Chinese first knew the use of the compass and had

Ch. 12]



applied it to navigation and that other traders, possibly Arabs, had carried the secret to Europe during the Crusades.20 The gradual evolution and westward movement of porcelain is better known. As far back as the Han dynasty (i.e., before a.d. 220) the Chinese had discovered that at a sufficiently high temperature a very fine glaze could be obtained with powdered felspathic rock mingled with lime.21 It is not, however, until about the seventh or eighth century that this glazed pottery was so perfected that it can be called porcelain. Its first appearance in the Near East was in the ninth century."- Porcelain manufacture was not known in Europe until after the Crusades. It is first mentioned in 1470 in Venice and the statement is made that the Venetians learned the art from the Arabs. It is in this world of varying and increasing currents of trade and intercourse that block printing started on its westward way. The trade that began under the wide empires of the Caesars and the Han and was furthered by the Caliphs and the T’ang, reached its culmination in the time of the Mongol Empire and the Crusades. Immediately after the Crusades new ideas of all sorts, some of which had their origin in the East, began to sweep over Europe. Whether or not the discovery of printing, that foundation stone on which modern education is built, is one of the gifts which Europe received from the East through the medium of the Mongol Empire and the Crusades, will be the subject of investigation in the next chapters— the discussion in Part Three being confined to block printing and in Part Four to the use of movable type.

NOTES 1. For early reports of discoveries of Neolithic culture in China and a discussion of its relation to Neolithic culture in other countries see J. G. Andersson, 1923a, b. The bibliography on this subject since 1923 is very rich. Andersson’s own final reports may be found in Bull, of the Museum of Far Eastern Antiquities 15 (1943), 17 (1945), and 19 (1947). 2. For full discussion of what Chang Ch’ien’s mission meant in the opening up of trade, see Laufer, 1919:535 ff. Chang Ch’ien found bamboo



[Ch. 12

staves and cloth from Szechuan already in use in Bactria, which he concludes came by way of India. His conclusion, of course, need not be taken at face value. Both bamboo and cloth may have come from India. 3. For additional details, and translations of the Chinese sources on which these statements are based, see Hirth, 1885. The contacts of Roman culture bearers with the estuary of the Mekong River have been confirmed by archae¬ ological discoveries made since 1944; cf. Coedes, 1947:193-99; and Malleret, I947~5o:h 189-99. One of a medal of gold represents the effigy of one of the Antonines and bears a mutilated legend in which the name of the emperor Marcus Aurelius (reigned a.d. 161-80) may be made out. Another medal carries the name of Antoninus Pius (Titus Aurelius, reigned 138-61) and indicates the fifteenth year of his reign, or 152. 4. The two authorities for the introduction of silkworm eggs during Justinian’s reign are Theophanes and Procopius. Theophanes refers to the country from which the silkworm eggs were introduced as the “land of Seres, Procopius as India. Both terms were at that time used very loosely by Greek writers. It is known from Chinese sources (one source is the narra¬ tion of the travels of Hsiian-tsang [602-64]; cf- Beal, 1906:11, 318-19; see also Hsin Tang shu 22iA/23b) that silk culture was introduced into Khotan in 419, and it is probable that it was from here that the eggs were introduced into Constantinople. 5. For fuller details and references to sources, see Beazley, 1897-1906:1, 186-91. Hennig-Dusseldorf, 1933:295-312, concludes that the monks men¬ tioned in Procopius secured the silkworms and learned their culture in Sogdiana, not in China, or, as is sometimes suggested, in Ceylon. The story has been most recently examined by Lieberman, 1953: chap. 6. 6. Lopez, 1952:72-73, reports that the earliest mention of the domestic production of raw silk in northern Italy (in the Po valley) dates from the first half of the tenth century; “a little later we hear of production in the island of Arbe (Rab, now Yugoslavia).” Spain and Sicily may also have produced raw silk about the same time. 7- Some efforts to improve the situation have been made during the last two decades, particularly by the Chinese. Cf. Goodrich, 1950b. 8. See Laufer, 1919:539-40. 9. The chicken (or its prototype) is indigenous to northern India and Burma. The date of its introduction into China cannot be determined; it was already known in the pre-Confucian period. It is first mentioned in Babylo¬ nian inscriptions in the seventh or sixth century, b.c. It is not mentioned in Homer or the Old Testament, but is constantly mentioned in the New Testa¬ ment. No bones of the domestic fowl have been reported from Shang or pre¬ historic sites in China proper, but Japanese archaeologists, who uncovered a prehistoric site near Port Arthur, characterized by both painted and black

Ch. 12]



pottery, assert that the prehistoric inhabitants of this area had fowl, pigs, and dogs as domestic animals. Cf. T. Kanazeki, et al., 1942:97-100. 10. The first mention of tea in Chinese literature, so far as is known, is in the biography of Wei Chao in the San \uo chih. Wei Chao died in a.d. 273 and the author of the San \uo chih in 297. Pelliot, 1922:432-34, contended that tea had not spread through North China until about the tenth century. This date seems too late. See the evidence collected in Goodrich and Wilbur, 1942:195-97. To those mentioned may be added one more. Tea drinking is frequently mentioned in connection with Po Chii-i (Waley, 1949:79, 117, 200), and Po’s cousin, Po Min-chung, who served as chief minister at the court at Ch’ang-an in 846, was called “Tea-boy.” Tea was very little known among the Mongols until the thirteenth or fourteenth century. On the other hand, tea is described by an Arab traveler in China in the ninth century (cf. Sauvaget, 1948:60), and its use apparently spread to Russia and Western Asia during Mongol times. In consequence of this the name for tea in Russian, Turkish, Persian, and Modern Greek is based on the north Chinese ch’a. It is not mentioned in European literature till 1588 when it was imported from South China by the Portuguese, whose tea trade was soon superseded by that of the Dutch. Hence the word “tea” and its variations, derived probably from the dialect of Fukien, is used in the languages of western Europe. 11. The carrot is apparently a native of northern Europe. It was cultivated by the Anglo-Saxons before they invaded Britain. It was carried by the Arabs into Persia in the tenth century. From there it entered China during the Mongol Empire (Laufer, 1919:451-54). 12. Several discoveries of glass beads of pre-Han date have been made. See especially White, 1934:30, 50. The dating is uncertain but it must be prior to the third century before our era, and may be as early as mid-sixth century b.c. The specimens of Chinese glass generally contain barium while Mediterranean glass does not. It would seem therefore that, while there is the strongest resemblance between Chinese and Mediterranean glass beads and the technique of making them may have been imported from West to East, the Chinese knew the art of manufacture. Cf. Ritchie, 1937:209-20 and Seligman and Beck, 1938:1-64. 13. All true alphabets in the world appear to have sprung from one early source in Phoenicia and Palestine. This alphabet reached India from its source by the Mediterranean almost as soon as it reached Greece. Through the early centuries of the Christian era the Indian forms of the alphabet vied with those coming directly from Syria for supremacy in Central Asia. The Tibetan alphabet, one of the Mongol alphabets, and the Korean alphabet were based on Sanskrit; while the alphabet of the Manchus goes back ultimately to a Syriac source, through Sogdian, Uigur, and Mongol as intermediaries. 14. The history of spinach and of sugar is also interesting as showing how ideas found their way in these early days through Asia and Europe. The earliest known reference to spinach in any literature is contained in the



[Ch. 12

annals of the T’ang dynasty, where it is stated that in the year 647 the king of Nepal sent some spinach to the Chinese Emperor T’ai Tsung. There is some evidence that Nepal got its spinach from Persia. At any rate it was in Persia that the Arabs found the vegetable not long after their conquest of the country. By the eleventh century it had spread through the Arabic dominions as far as Spain, but it was not until the fourteenth or fifteenth century—after the Crusades—that its entrance into Christendom is recorded. (Laufer, 1919: 392-98.) Sugarcane was imported into China as early as a.d. 285 from Indochina and was again imported into China from Persia during the seventh century. During the seventh century also a special mission was sent by the Chinese emperor to Magadha in India to learn the process of boiling sugar, and this Indian method was adopted by the sugar-growers of Yang-chou. Through the Middle Ages the Saracen empire was the center of sugar production. Sugar¬ cane was introduced by the Arabs from Persia into Egypt, Sicily, and the south of Spain. As late as the thirteenth century sugar refiners from Cairo came to China to teach the superior methods of sugar refining that were practised in Egypt. From Cyprus and Sicily sugar production was carried to Madeira about 1420 and to the Canaries in 1503. Sugar production in Brazil and Haiti began also very soon after the discovery of America. To this digest of the remarks of Laufer, 1919:376-77, may be added one later reference: Lippmann, 1929, esp. Chapter 7. He considers that the poet Ssu-ma Hsiang-ju (d. 117 b.c.) refers to sugarcane. Other examples of the spread of ideas through the Euro-Asiatic continent are found in the study of such games as dice, chess, and backgammon. The spread of playing cards, as more closely related to printing, is studied in Chapter 19. 15. Yahb-allaha III, patriarch of the whole Nestorian church, with seat at Baghdad, 1281-1317. For a discussion of the organization of the Nestorian church in Asia during the Middle Ages see Dauvillier, 1948:261-316. 16. This place, rendered Hanfu in Arabic, has sometimes been identified with Hangchow. For its identification with Canton, see Pelliot, 1922:410, and Kuwabara, 1928:10-11. The correct date of the rebellion is 879, according to Chinese records. 17. See Ferrand, 1922:85-92. 18. Goodrich and Feng, 1946:114-23, 250-52; Wang Ling, 1947:160-78; and Feng Chia-sheng, 1947:6, 29-84; and 1949:1-51. On Arabic acquaintance with and names for saltpeter and rocket see Laufer, 1917:74 and Ley, 1941: 14. For Roger Bacon’s cryptic announcement of gunpowder of 1249 see Montross, 1944:176-77. 19. On the “south pointing chariot” see Wang Chen-to, 1937:1-47; Lanchester, 1954:1-4; Moule, 1924:83-98; and Hashimoto, 1926:69-83. The latest study is by Li Shu-hua, 1954:78-94, 175-96.

Ch. 12]



20. Klaproth, 1834; Hashimoto, 1926:83-92; Kawabara, 1928:69; and Ferrand, 1928:31-127. The last includes the long essay on the subject (with corrections by Pelliot and Ferrand) of Leopold de Saussure, 1923. 21. Fujio Koyama, 1941:383-89, pushes the manufacture of real porcelain back to the end of the first century


22. Cf. Hobson, 1936-37:11-17, especially p. 17; and Kahle, 1940-41:32-33.

Chapter 13 PAPER’S THOUSAND-YEAR JOURNEY FROM CHINA TO EUROPE Paper has everywhere been the forerunner of printing.


this strong economical material, printing could never have made headway. Moreover the westward movement of paper not only pre¬ pared the way for printing, but its history is often suggestive of the ways in which printing may have traveled. In order to investigate the course of block printing, it is therefore necessary to understand clearly the history of paper. This history of paper is open before us. As compared with that of block printing, the advance of paper was a triumphal progress, hailed by literary men, and displacing quickly the old writing materials in every place it touched. Typography later met with a like welcome, first in Korea, then in Europe. But block printing was always in its beginnings an obscure and despised art, whose history can be traced only with the greatest difficulty. A study therefore of the progress of paper affords the best introduc¬ tion to the more difficult study of the westward course of the block printing which followed in its wake. The invention of paper from hemp, tree bark, fish nets, and rags, as officially announced to the emperor of China in the year a.d. 105, has been described in Chapter 1. The history of the later Han dynasty, compiled by Fan Yeh (398-445), states, “From this time on it was used universally.” Other references confirm the impres¬ sion that its spread through China was very rapid. West China is especially noted by several writers as one of the early seats of the paper industry.1 The first point at which paper is met on the journey from China westwards is at a spot near Kharakhoto (in the modern province of Ning-hsia) together with many wooden documents dated between 132

Ch. 13]

the years



93 and 98. Conceivably, therefore, this fragment might

belong to the decade in which Ts’ai Lun’s announcement was made. The next is at Loulan, found by Sven Hedin’s expedition, and is believed to date from a little after


250.2 Each of the older sites

excavated in Turkestan yields both wood and paper as writing material. At several places the gradual displacement of wood by paper in the third, fourth, and fifth centuries can be traced. At Loulan, for example, abandoned about 350, some 20 per cent of the many documents found by Sir Aurel Stein are on paper, the balance on wood. For a time it was believed that certain other letters found by him in a ruined watchtower of a western spur of the Great Wall, near Tun-huang, amid a mass of documents written between 98 and



153 on wood and silk, might be earlier than those discovered

by Hedin. It has recently been shown, however, that they are of later origin. Dr. Henning has tentatively set the date of these letters, all written in Sogdian, at 312-13.3 At Turfan the oldest paper found by the Prussian expeditions dates from 399. Here paper coming from the east met the culture currents that were coming from the west and south. From here we have early Aramaic texts on paper, and even some three or four words of Greek. Also written on paper is a fragment of a Bible manuscript (from the Book of Psalms) in Persian, which has been dated by some as early as 450. The early paper from Turfan in¬ cludes Manichean texts, Buddhist canons, and a variety of Christian literature—among other things a delightful fairy story based on the visit of the Wise Men to Bethlehem. Step by step, paper penetrated around both edges of the Takla Makan desert, till by the end of the fifth century, through all the Central Asian territory, which then was under Chinese control, ex¬ cept in certain backward spots, the use of wood for writing had stopped and paper was in general use.4 In the early years of the eighth century the Arabs mastered what is now known as Russian Turkestan. Already paper had entered this region somewhere between the fourth and sixth centuries and was in occasional use.5 So far as is known it was all imported from China. In July, 751, however, paper manufacture entered the Arabic



[Ch. 13

world and started on its career from Samarkand to Spain. The circumstances are related in detail in the Arabic annals. There was war between two Turkish chiefs. One appealed for help to China, the other to the Arabs. The Arabs defeated the Chinese army and drove them back as far as the Chinese frontier. Among the prisoners taken were some papermakers, who taught the art of papermaking at Samarkand.6 The Chinese annals of the Tang dynasty describe the same battle and the date exactly agrees.' The Arabic report states that the paper introduced into Sam¬ arkand was made of “grasses and plants.”

On the other hand,

all early Arabic paper that has come down to us, including the great Rainer collection, is rag paper. An examination of a number of papers of just this period (768-87) from Eastern Turkestan shows that they are made of a mixture of rags and raw fibers, the raw fibers predominating.8 The Arabs seem to have found difficulty in getting all the materials that had been used by the Chinese and made their paper wholly of rags, like the earlier Chinese paper found by Stein in the Great Wall. “Paper of Samarkand” soon became well known through the Asiatic dominions of the Caliphate—so much so that a century later (869) Juhith wrote, “The papyrus of Egypt is for the West what the paper of Samarkand is for the East.” A writer of the eleventh century, Tha’alibi, writes: Among the specialties of Samarkand that should be mentioned is paper. It has replaced the rolls of Egyptian papyrus and the parch¬ ment which were formerly used for writing, because it is more beau¬ tiful, more agreeable, and more convenient. It is found only here and in China. The author of the work “Journeys and Kingdoms” tells us that paper was brought from China to Samarkand by pris¬ oners. It was Ziyad, son of Salih, who took those prisoners, among whom were found the papermakers. Then the manufacture grew and not only filled the local demand but also became for the people of Samarkand an important article of commerce. Thus it came to minister to the needs and well-being of mankind in all the countries of the earth.

But long before this a rival factory had started at Baghdad. In the year 793-94 Harun-al-Rashid of Arabian Nights fame brought

Ch. 13]



Chinese workmen for the starting of the first paper factory in the capital. Baghdad, however, does not seem to have seriously rivaled Samarkand as a source of supply. Already in the tenth century we find Arabic scholars debating with warmth whether the Ommiads or the Abbassids had the honor of the introduction of paper at Samarkand, the change of dynasty having occurred in 750. The third factory recorded in the Arabic empire was on the southeast coast of Arabia. The fourth was Damascus, which was for several centuries the main source of supply for Europe, paper in Europe being generally known as charta damascena. Mambij or Bambyx, another Syrian town, seems also to have given its name to paper in Europe, with strange consequences. For charta bambycina, paper of Bambyx, was corrupted to charta bombycina, paper of cotton, and from the time of Marco Polo down to 1885, when the view was disproved by the microscopic investigations of Dr. Wiesner, Arabic paper and early European paper were known as “cotton paper” and the invention of rag paper was attributed to the Germans and Italians of the fifteenth or sixteenth century.9 But while Damascus became the center of the export of paper to Europe, the secret of its manufacture was destined to enter Christendom by a longer route. It had come the full length of Asia; it was yet to go the full length of North Africa and enter Europe through Spain. Paper in Egypt has special interest because it is here in the desert that almost all the old Arabic paper has been found—just as the desert of Turkestan has been the storehouse for old Chinese paper. The Erzherzog Rainer collection at Vienna contains about 12,500 documents, mostly written on paper, dating from about 800 10 to 1388. It is the examination of this Egyptian paper which has thrown the greatest light on the history of paper manufacture. The steady displacement of papyrus by paper is interestingly illustrated by the dated documents in this collection. Of the second century of the Hegira (719-815) there are thirty-six dated docu¬ ments all on papyrus. From the following century (816-912) there are ninety-six documents on papyrus and twenty-four on paper.



[Ch. 13

From the fourth century (913-1009) only nine are on papyrus and seventy-seven on paper. The last papyrus dates from 936. A polite letter of thanks, whose date lies between 883 and 895, closes with the words, “Pardon the papyrus.” As the letter is written on a most beautiful piece of papyrus, the writer is evidently apologiz¬ ing for not using paper, which although just introduced from Bagh¬ dad or Samarkand was already the stylish material for letters. A Persian traveler, writing about 1040, recorded with surprise how in Cairo “the venders of vegetables and spices are furnished with paper in which everything that they sell is wrapped.”


physician from Baghdad, writing a century later, reveals the source of this wrapping paper used by the grocers: “The Bedouins and fellahs search the ancient cities of the dead to recover the cloth bands in which the mummies are swathed, and when these cannot be used for their clothes, they sell them to the factories, which make of them paper destined for the food markets.” Mercifully no paper mill was set up to use the textiles in the tomb of Tutankhamen! From Egypt paper manufacture passed to Morocco and thence to Spain. The first clear mention of the making of paper in Spain —which is also the first in Europe 11—indicates a well-established industry. It was in 1150 diat El-Edrisi said of the city of Xativa, “Paper is there manufactured, such as cannot be found anywhere in the civilized world, and is sent to the East and to the West.” For a century still, the paper manufacture of Spain was alto¬ gether in Saracen hands, though Christians seem gradually to have learned the art as the Christian conquest advanced. The first re¬ corded paper mill in Christendom was founded in 1157 by Jean Montgolfier 1J in Vidalon, not far from Herault on the French side of the Pyrenees, though for still another century Europe’s paper was largely supplied by the Saracen mills of Damascus and Spain. So for its first six hundred years papermaking was a Chinese monopoly, until taught to the conquering Arabs by Chinese pris¬ oners at Samarkand. For the next five hundred years papermaking in the West was an Arab monopoly until the Arabs in turn taught their Christian conquerors in Spain, and Christendom made ready to take the lead.




Meanwhile paper was entering Europe by two other routes. Paper of Damascus was becoming a large article of commerce, chiefly through Constantinople, and paper from Africa was enter¬ ing through Sicily. As there were paper mills in southern France in the thirteenth century, paper, either of Spanish or Arabic origin, may also have come in through this route. The earliest extant paper document from Europe comes from Sicily. It is a deed of King Roger, written in Arabic and Latin, and dated 1109. A manuscript on paper, a part of which dates from IX54>

still preserved in the archives of Genoa. Emperor Frederick

II in 1221 forbade the use of paper for public documents, but the prohibition was not altogether effective. The import into Italy of paper from Damascus increased steadily through the thirteenth cen¬ tury, and by 1276 the first Italian paper factory had been set up at Montefano. Italian paper manufacture spread rapidly, and Italy in the fourteenth century soon rivaled and then outstripped Spain and Damascus as the source of Europe’s supply. In Germany the use of paper increased steadily during the four¬ teenth century, especially during the latter half, but all paper was imported—largely from Italy. Toward the end of the century, when block printing first appeared, South Germany was buying its paper supply from Venice and Milan, and the Rhineland from France, though import from Damascus had not altogether ceased. The use of the new writing material was just beginning to be general. Its employment was not yet as common as that of parchment. Nurem¬ berg, which was one of the earliest centers—perhaps the birthplace— of the block printer’s art, has also the honor of being the first place in Germany, so far as is known, where paper was made.13 This first paper mill was started in 1390. The date of the earliest block print¬ ing is uncertain, but it was probably at just about the same time.14 The slow advance of paper manufacture in Europe, which can readily be seen from a glance at the chart on page 245, is in startling contrast to the very rapid advance of printing when it once started on European soil. Paper seems to have advanced less rapidly in Europe than it had advanced either in China or in the Arabic world. The European parchment with which paper had to compete was a



[Ch. 13

far better writing material than either bamboo slips or papyrus. Furthermore, there were few in Europe who read, and the demand for a cheaper writing material, until the advent of printing, was small. While it was the coming of paper that made the invention of printing possible, it was the invention of printing that made the use of paper general. After Europe began to print, first from blocks and then from type, paper quickly took its place as the one material for writing as well as for printing, though, strange to say, the first paper mill in England was not set up until seventeen years after Caxton began to print at Westminster.

NOTES 1. Blanchet, 1900:16-17. Note especially quotations from Chou tien p’u, Tien chen p’u, and Ten ts’ao \ang mu. 2. This paper from Loulan, containing a fragment from the Classics, is undated, the date being estimated from the style of writing, etc. The three earliest dated manuscripts from Loulan are of the years a.d. 252, 266, and 310,* according to Clapperton, 1934:9. Of a fourth manuscript he writes: “Another of the documents found at the same time is supposed to date from the second century.” Folke Bergman is not convinced by the arguments for the early dating of this manuscript, known as Hedin document 1, and is inclined to put it later. He also omits completely the date of a.d. 252 from his list of dated documents from Loulan. See Bergman, 1935:76-78. Maspero, 1953:52, reports that certain paper documents of the third expedition of Stein also date from the years 263-70 to 312-30. This is likewise true of some found by the Otani mission. 3. The original dating, questioned by Pelliot, 1931:458-59, has apparently been settled by the study of the contents of the letters; see Henning, 1948: 601-15. 4. There are one or two isolated points where the use of wood for writing persisted parallel with that of paper. At Miran it continued until about the eighth century. Otherwise the triumph of paper by the end of the fifth cen¬ tury was complete. In India, where cloth was commonly used for writing material, paper became known in the seventh century but was not manufactured until about the year 1000. Cf. Gode, 1944:87; and 1952:1-7. See also Bagchi, 1937:351—52, who demonstrates that the eighth century Sanskrit word for paper, $aya, is based on the Chinese chih, then pronounced tsie according to Karlgren, 1923: No. 879.

Ch. 13]



5. Pelliot, in an undated five-page leaflet “Explorations et voyages dans la Haute Asie, published about 1931, hazards the guess that a collection of Sanskrit texts on paper found by the Citroen expedition at Gilghit in the Kashmir region may go back to the fourth century. Sylvain Levi more con¬ servatively placed them in the sixth century. See Blum, 1934:54. R. N. Frye, 1951:123, draws attention to Chinese paper found on Mt. Mugh, 120 kilometers east of Samarkand, dating from ca. 709-23. A. von Le Coq, 1923:16, advances the suggestion that the Sogdians may have been largely responsible for the diffusion of the techniques of papermaking. They manufactured paper in large amounts both in Sogdiana and in Eastern Tur¬ kestan. The Manicheans as well as the Buddhists were users of paper. 6. Hoernle, 1903:663 ff. 7. Hirth, 1890:270; Chavannes, 1900:297. Besides the prisoners taken in 75L Chinese were also captured in 775— 85 by the governor of Samarkand. Tamlm Ibn Bahr, writing of this around 821, adds that they “fabricate in Samarkand good paper and various kinds of arms and instruments.” See Minorsky, 1948:285. 8. The raw fibers are largely those of paper mulberry, laurel, and China grass (Boehmeria nivea). 9. The following, written in Parma in 1782 by Andrez, and quoted by Thomas, 18x0:37-38, indicates one of the views that used to be held with regard to cotton paper, “Paper made from silk was anciently fabricated in China, the art of making this paper was carried from China to Persia about the year 652 and to Afecca in 706. The Arabs substituted cotton and carried the art of papermaking into Africa and Spain.” 10. The earliest document on paper is dated a.d. 874. “Two other undated letters are believed to belong to a.d. 791.” Clapperton, 1934:7, 61. 11. The famous manuscript of the Convent of San Gilos, dating from 1129, has alternate pages of parchment and paper. This may have been paper imported from Africa, though it is more likely that it was Spanish, and would therefore antedate the statement here quoted from El-Edrisi. 12. He learned how to manufacture paper at Damascus while a prisoner of the Saracens (Fuhrmann, 1938:30). Hunter, 1947:470-73, gives the following dates for the appearance of paper in various countries of Europe: Spain 950, Constantinople 1100, Sicily 1109, Italy 1154, Germany 1228, England 1309, Holland 1346 (1322?). 13. Cologne and Mainz both claim to have had paper factories as early as 1320, but the claim is disputed. Nuremberg’s manufacture of paper is the first that is known with certainty. 14. It seems fitting that Nuremberg should have been the home a century later of Albrecht Diirer, who was not only Germany’s greatest painter, but also a maker of woodcuts on paper.


The important

position held by China’s western borderland in the

early history of block printing has already been noted. Some of the earliest literary references and the oldest prints that have been found come from the two far western provinces of Szechuan and Kansu, while printed books of the eleventh century have even been found across the border in Mongolia. Though the better preservation of documents in such places as Tun-huang and Kharakhoto is chiefly due to the climate, yet one may assume also that most of the earliest printing centers were in the western part of China, and that from this part of the country the new art spread not only eastward but westward. The region where the greatest quantity and the greatest variety of early block prints has been found is the oasis of Turfan in what is now Chinese Turkestan. This oasis of Turfan, some four hundred miles northwest of Tun-huang, is a strange depression in the earth’s surface almost as deep as that of the Dead Sea, and surrounded on all sides except that toward China by high mountains. A people of Indo-European speech developed the first recorded civilization in the Turfan basin, going back to a period before the Christian Era. In the early centu¬ ries after Christ, Buddhism swept through the region, bringing in its train a highly developed literature and art. Manichean and Chris¬ tian (Nestorian) missionaries began coming to the Turfan oasis about the fifth or sixth century and their influence soon rivaled, though it never displaced, that of the Buddhists, bringing into the country a considerable element of Persian and even Byzantine cul¬ ture. In the seventh century the oasis was conquered by the Chinese and from that time on down to Mongol times Turfan was intermit¬ tently under Chinese control, though the hand of the Chinese over140

Ch. 14]



lords was rather lightly felt, and the country was left to develop its own peculiar institutions. More important was the conquest by a powerful Turkish tribe called the Uigurs in the eighth and ninth centuries, for these Uigurs soon made Idiqut, near the modern Turfan, the capital of their empire, and adopted as their own the older civilization that they found. From this time Turfan, located as it is in the very center of Asia, may be said to have been a focal point where culture streams of all Asia met, open on the south and west to the religious influences of India, Persia, and Syria, open on the east to the political hegemony of China, and on the north forming the cultural center of a loose empire that stretched far away over the nomad tribes of Mongolia and even Siberia. The Uigur civilization came to its height in the ninth and tenth centuries of our era, but Turfan remained an im¬ portant cultural center until after its conquest by the Mongols under Jenghis Khan. This Turfan basin was first excavated by the Prussian expedi¬ tions of Dr. Griinwedel and Dr. von Le Coq in the years 1902 to I9°7- The results of these excavations, including a large quantity of woodcuts and block prints, were in the Ethnological Museum at Berlin until World War II, when they suffered loss by bombing. It is the mingling of races and religions that gives to the Turfan discoveries their peculiar fascination. Chinese and swarthy Indians, Turks and blue-eyed, fair-haired mountaineers of Tocharian race, all stand out clearly in the wonderful wall frescoes, while the man¬ ner of portrayal is a blending of the art of Greco-Indian Gandhara with that of China, not to speak of considerable Persian or Iranian influence. Nor is there less mingling in the domain of religion. Side by side stood the churches of the Christians and the temples where Buddhists and Manicheans seem to have worshipped together.1 All three religions flourished during the period before the Turkish con¬ quest, and the conquest of the country by the Uigur Turks brought little change, except that Manicheism became the religion of the reigning house. Christianity was apparently always tolerated and Buddhism encouraged by the new overlords. Islam, though sweep-




ing over the lands directly to the west, left the older religions in the Turfan basin undisturbed until after Mongol times. Manicheism was the religion of the royal house, Buddhism that of the majority of the people, Nestorian Christianity that of the minority. The Confucian culture of the Chinese overlords made little impression. Needless to say, Turfan was a polyglot community. Seventeen different languages are represented among the documents found by the Prussian expeditions, including Syriac, Persian, Sanskrit, Chi¬ nese, and a few words of Greek, as well as the local Tocharian and Turkish. Some of these appear in as many as four or even five dif¬ ferent alphabets. There seems to have been a mania for fitting new alphabets to the various languages of the oasis. The religious impulse motivated all this literary, linguistic, and art activity. In the ruined monasteries and temples the ancient docu¬ ments have been preserved: Christian scriptures, Manichean hymns and prayers, and Buddhist sutras that form the bulk of all the manu¬ scripts found. In this melting pot of race, language, and religion, with its high valuation of literature and art, block printing early found itself at home. It has already been pointed out how Buddhism had a par¬ ticular genius for reduplication, and some of the earlier, cruder manifestations of that genius as found in the Turfan monasteries have been described. It is a significant fact that all the block print¬ ing of Turfan so far found is Buddhist. Woodcuts and block prints were found in almost every site excavated in the Turfan region. Toqsun at the western edge of the Turfan oasis is the most western point at which Central Asiatic block printing has been discovered. The state of preservation of the Turfan texts is very different from that at Tun-huang. In contrast to the neatly piled rolls of an undisturbed sealed chamber, the Turfan manuscript treasures show signs not only of the natural destruction of the centuries, but of wanton destruction as well.

In certain of these monasteries the

floors were littered with papers, all either hacked to pieces, or else crushed by hand, piece by piece. In the monastery at Idiqut, for instance, the floor was covered knee deep with this “waste paper.” One could have carried away many hundred pounds of it. In the

Ch. 14]



midst of this deposit were the corpses of several Buddhist priests, evidently killed while the systematic destruction of their library was going on. Manichean documents were mingled with those of Bud¬ dhist origin, but, whether by design or by accident, they do not seem to have been so thoroughly destroyed. One box of this deposit examined at Berlin—a box that had remained packed away ever since it came from Turfan and from all appearances might have been the contents of a wastepaper basket in a modern Chinese school¬ room, plus a generous accumulation of dust—contained, among crumpled and torn manuscripts in Uigur, Sogdian, Chinese, and Sanskrit, a dozen very primitive Buddhist woodcuts, two printed texts in Uigur, a sheet of stamped Buddhas (hand-colored), a Chi¬ nese manuscript with a stamped Buddha at the top of each column, several bits of silk with Buddhist figures stencilled upon them, and a bit of printed silk. Most of the printing of the Turfan region has had to be rescued from such crumpled deposits.

But one notable exception is the

monastery at Murtuk, in which a large proportion of the best block prints were found. This monastery seems to be of later date than most of the others. Perhaps its documents were produced after the persecuting zeal that destroyed the other libraries had spent itself. Murtuk as compared with other sites is remarkable for three things— a larger proportion of its documents is printed, its printing is better done, and its printing is far better preserved. Of all the printed documents found in the Turfan region, not one is dated. Nor is the approximate dating easy, especially of the earlier pieces. With regard to the later ones, it is possible to speak with more clearness. There are four fragments in the Mongol lan¬ guage, also a beautiful large Sanskrit book in Lantsa script, and a fragment containing the name of Jenghis Khan—-all of which could not be earlier than the opening decades of the thirteenth century. Moreover, as it is known that the Uigur civilization did not long survive the drain on its man power caused by the Mongol wars, the date at which the Turfan documents come to an end cannot be much later than the close of the thirteenth century. It may there¬ fore be said with a fair degree of certainty that a number of the



[Ch. 14

best printed pieces—and perhaps a very considerable number—be¬ long to the thirteenth century and the opening years of the four¬ teenth, when Uigur printing came to its climax and ended. How far back of this the art goes can be only matter for conjecture. It may possibly go back as far as that at Tun-huang or further. It is certain that there is a large amount of very primitive printing and near-printing, which may indicate several centuries of development. Some would assign much of the printing in the Uigur language to an early date, because the Uigur civilization rose to its height during the ninth and tenth centuries. But all this is conjecture. Whatever may be the date at which Uigur printing began, there seems to have been continued progress both in quantity production and in quality. Late monasteries like Murtuk are much richer in block prints than the earlier ones, and the printing is better. It may then safely be said that there was during early Mongol times in the monasteries of the Turfan region a highly developed and widely extended print¬ ing industry, which had very likely been going on for several cen¬ turies. Six languages are used in the Turfan block prints: Uigur, Chi¬ nese, Sanskrit, Tangut, Tibetan, and Mongol—the Uigur, Chinese, and Sanskrit prints predominating.2 The Uigur books and fragments are in the Sogdian alphabet— an adaptation of Syriac that had penetrated into Central Asia early in the Christian Era and had been taken over from the people of Indo-European speech by the Uigur conquerors. The language is a pure Turkic, which, though not the direct ancestor of modern Turkish, presents a striking likeness to it. As the books are all translations of Buddhist sutras, they contain many transliterations of Sanskrit names and words. Where this occurs, the Sanskrit orig¬ inal is printed in between the lines, much as English words are introduced in a modern Japanese text. The page numbers in Uigur books are as a rule in Chinese, as is also the title of the book, which appears at the side of many of the pages. These Buddhist printed books in a Turkish language with Sanskrit notes and Chinese page numbers, in a script brought from Syria, are in themselves an epit¬ ome of the eclectic character of the Uigur civilization.

Ch. i4]



The Chinese books, of which there are many, are as a rule excel¬ lently printed in the large bold-faced style characteristic of the Sung era. They are better printed and easier to read than any modern Chinese block books. The Chinese books, like those in Uigur, are in the main translations of Sanskrit sutras. Both the Uigur and the Chinese books are usually in the folded form, but there are also a few printed rolls, which may very likely indicate an early date. The Sanskrit prints are of two kinds. The larger number are in an older script, which shows little change from that which was already in use in Central Asia several centuries before block printing began. A few Sanskrit prints are in the later Lantsa script, which was not in use before Mongol times. The most beautiful specimen of printing in the entire collection is a Sanskrit Diamond Sutra in Lantsa script. Each page—some ten leaves have been found—is more than two feet long by six inches wide, with broad margins and beautiful clear print. The titles and page numbers on alternate pages are Sanskrit and Chinese. This Sanskrit edition is later than the Chinese editions of the same book found at Tun-huang, at Kharakhoto, and in Japan; and, judging from the script, it may be said with a fair degree of certainty to date from the thirteenth century. Apparently printed on both sides of the page, each sheet actually consists of two leaves pasted together with such nicety that the past¬ ing can scarcely be detected. This, like many of the Sanskrit printed books, retains the Indian form rather than the Chinese. It is a pothi', that is, it is like the ancient books that were written on palm leaf in India, many long narrow sheets laid between two boards and bound through with a thong. The printed books of Turfan afford an interesting study in the competition that was going on between different book forms—the roll, which was the earlier form both in China and in the West; the folded book, which under the influence of printing had gradu¬ ally displaced the roll in China; and the Indian pothi. The one form lacking is the stitched book, familiar in the West. This omis¬ sion is interesting, as Christian and Manichean stitched books had begun to circulate in the Turfan region not so very long after their



[Ch. 14

first use in Syria in the fifth century. The stitched book reached China early in the Sung dynasty (about the eleventh century) and most of the printed books of that period from China that are now extant are stitched. Somehow the Buddhist has never taken kindly to this form. In Turfan it was used in the main by Christians and Manicheans; in China it was the mark of Confucian and secular literature, and it came in time to be the usual form in the West. The Buddhists always preferred the folded book—or sometimes in Cen¬ tral Asia the Indian pothi. In fact a curious form frequently met among the Turfan printed books is a cross between the two—a folded book copied exactly from a manuscript pothi, with the old holes for the pothi thong copied in the printing. The Tangut printing is not extensive. It is in a script only par¬ tially deciphered, ideographic, and evidently based on Chinese, yet differing radically from Chinese—one of man’s very few attempts within relatively modern times to create an ideographic script. It was the language of a powerful kingdom—racially akin to the Tibetans—which held sway in Kansu and adjacent territory during the two centuries before the Mongol conquest. This Tangut print¬ ing can therefore be dated with a fair degree of accuracy.3 The Tibetan prints, though not the oldest, are the crudest of those found at Turfan. They are mostly charms of two or three words each, contained in little clay Buddhas, which have to be broken in order to remove the printed charm. There are just four fragments of Mongol printing. They are bits of sutras, printed in the ’Phags-pa script that was derived from Tibetan, and not in the more usual script that the Mongols took over from the Uigurs. The Turfan finds include also a large number of woodcuts and fragments of woodcuts, without text. These are apt to be on very thin paper and rather primitive in workmanship, though there are notable exceptions. The discovery at Tun-huang of a font of Uigur wooden type be¬ longing to early Mongol times naturally arouses the question whether any of the Turfan printing, especially that in the Uigur language, could have been done with movable type. This Uigur

Ch. 14]



type will be more fully discussed in Chapter 22. All that can be said here is that there is no evidence of the use of type at Turfan. Nor is there evidence to the contrary. The difference between a blockprinted book and one printed from such type as that found at Tunhuang would be difficult to detect. The part played in the spread of printing by peoples of Turkish extraction is an interesting study. The tenth century was a great century for the Turks. During parts of this century, while the Turkish civilization of the Uigurs of Turfan was at its height, Turks of other tribes were ruling China, Egypt, and the Baghdad Caliphate. This vast territory under Turkish rule, stretching from the Pacific to beyond the Nile, did not in any sense constitute a single empire, and it is doubtful whether the Turkish emperors on the throne of China were even aware that men of their own race were ruling in Cairo and Baghdad. Yet it was Turkish individuals—adventurers— who had seized the power in all three lands and it was with Turkish armies that they held power. The founders of the three short-lived dynasties that ruled China from 923 to 951 were, like their con¬ temporaries in the Moslem empires, Turkish mercenaries who be¬ came sufficiently strong to usurp power. The birthplace of these adventurers in China was in the region of Hami, not far from Tur¬ fan. The home of the rulers of Egypt and Mesopotamia was a thousand miles or so to the west across the mountains. Yet in lan¬ guage and racial affinity they were closely related. The fact that the tenth century was a time in which block print¬ ing made such progress—the century of Feng Tao in China, of most of the block prints of Tun-huang, and possibly of the earliest prints both of Turfan and of Egypt—brings up the question whether there is any connection between the spread of block printing and the spread of the various branches of the Turkish race—an interest¬ ing subject for further study. The theory has even been advanced that block printing was primarily a Uigur or Central Asiatic in¬ vention.4 But the little Chinese page numbers on all the Turfan books, whether the language is Chinese, Sanskrit, or Uigur, are a sure indication of Chinese workmanship.

Block printing comes

from China. The fact that a larger number of early prints have



[Ch. 14

been found in Tun-huang and Turfan than in China proper is due to the climate. The great significance of the printing of the Uigur Turks lies in the fact that the Uigur civilization was taken over in toto by the new Mongol empire. The conquest of the Uigur realm


1206) was one of the first important achievements of Jenghis. From that time not only did the Uigurs form a large part of the Mongol army—they were also the Mongol brains. It was Uigurs who reduced the Mongol language to writing and applied to it their own alpha¬ bet. It was Uigurs who did such writing as was needed at the Mon¬ gol court. A Uigur was appointed by Jenghis as tutor to his sons “to instruct them in the language, laws, and customs of the Uigurs.” c Under Jenghis’ grandsons, the accountants and chief officers of state in Persia and in Mesopotamia were Uigurs. As Turfan, drained of its manpower for the Mongol armies, dwindled in importance, its culture was transferred bodily to Karakorum, and became the basis of such culture as the Mongols possessed, until it was gradually dis¬ placed at the eastward end of the empire by the higher civilization of China and at the westward end by that of Islam. During the lightning campaigns of the Mongols that resulted in the conquest of China, Persia, Mesopotamia, and Russia, it was the culture of the Uigur Turks that followed the Mongol arms. And the Uigur Turks were a people who knew well how to print.

NOTES 1. This mingling is illustrated in the Buddhist monastery of Toyok near Turfan, which now through a curious confusion of faiths has become a point of pilgrimage for Moslems from India and Arabia. Among the papers found in this old monastery were an enormous number of Chinese Buddhist manu¬ scripts, fragments of Indian manuscripts written on birch bark and on palm leaf, fragments in a still unknown Semitic script, several manuscripts in old Turkish runes, some Sogdian writings, Manichean writings in Turkish and Persian, Uigur writings in four different kinds of script, some Syriac frag¬ ments, some Manichean and Buddhist embroideries, and some beautiful Manichean miniatures. In the ruin of an old church near by are a large

Ch. 14 j



number of Christian texts in Syriac. No wonder the Moslems regard as sacred the place where so many peoples and faiths have met! 2. It is interesting to note that these six languages are exactly the same as those which Pelliot found in manuscript and printed remains in one of the later caves at Tun-huang. Furthermore a stone found at Tun-huang, dated contains parallel inscriptions in these same six languages. To this list may be added the well-known hexaglot inscription of 1345 in the archway at Chii-yung-kuan, a pass in the Great Wall north of Peking. Four of these languages (all except Mongol and Sanskrit) are mentioned by de Rubruquis as the languages used for writing at the Mongol court when he visited the Grand Khan in the middle of the thirteenth century. 3. The Tangut script was officially adopted in 1036; cf. Baruch, 1933: 22-24. F°r brief description and historical sketch of the Tangut, see Cordier, 1920:!!, 199-203. 4. “The Uigurs, an ancient Tatar people, have at all times been celebrated in Tatary. They have cultivated sciences and arts. . . . They write like the Chinese from the top down; and were the first to use wooden blocks for printing.” De Guignes, 1756-58:1, 7. Strange to say, this was written in the eighteenth century, before the discovery of any of the Uigur printing at Turfan. 5. Skrine and Ross, 1899:155-57.

Chapter 13 ISLAM AS A BARRIER TO PRINTING hundred years before block printing came into Europe, all East Asia was printing—from Nara to Turfan, fapanese, Koreans, Chinese, and Uigur Turks—and through most of this territory print¬ ing was being carried on on a large scale. But between the Far East that printed, and Europe where printing was unknown, lay the Mos¬ lem world that refused to put its literature in printed form. This bar¬ rier between the Far East, where all Buddhist, Taoist, and Confucian literature was being spread abroad in printed form, and Europe where ancient manuscripts were being so laboriously copied by hand in the Christian monasteries, proved in the end to be not impene¬ trable, but for a time the isolation of Europe from the lands of the

For several

Far East was complete. It is strange that such a literary, and such a religious, people as the Arabs refused to use this vehicle for the spreading abroad of religious thought. Paper they found in Central Asia—and with almost incredible quickness it displaced all other writing materials from Samarkand to Spain. But not so with printing. The reason for this prejudice is uncertain. It has been suggested that the Mos¬ lem suspected hog’s bristles in the brush used for cleaning the block, and that to touch the name of Allah with this brush seemed to him the height of blasphemy. It is more probable that mere conserva¬ tism was back of the prejudice. The Koran was given in written form, therefore the Koran must always be written. Whatever the reason may be, up to today the Koran has never been printed in any Moslem country except by block printing or lithography. (But note the exception in Chapter 18.) In 1727, when permission was asked by a Hungarian by the name of Ibrahim for the erection of a printing press at Constantinople, the Ulema under Sultan Ahmed III delivered 150

Ch. 15]



a verdict that it was against the religion and honor of Islam to allow the printing of the Koran, because the Koran rested upon written tradition, and must in no other way be handed down. Permission to set up a press was finally given him on condition that the Koran should not be printed, and in 1729 1 a history of Egypt appeared. But it awakened such opposition that until the nineteenth century no more printing was attempted in Moslem lands, and even through the nineteenth century printing has had to fight against great odds. There was printing done in Syria in the sixteenth century by Syrian Christians. Printing had been done in Arabic in Italy before the end of the fifteenth century," and the Koran was printed at that time. Catherine II had the Koran block printed in Russia in 1787. But, so far as is known, with the exception of the abortive projects of 1714 and 1729 at Constantinople, the Islamic world (the Chinese part of it excepted) never printed a book till 1825, when the first press was set up in Cairo. In China, at least twenty books were printed be¬ tween 1657 and 1825.3 During all the early period of Chinese printing the Arabic world was in close touch both with China and with Chinese Turkestan, and before the period was over, intelligent Moslems could not have been wholly ignorant of the role that was being played by literary and religious printing in the lands to the east. The growth of in¬ tercourse across Asia during the Tang dynasty has already been sketched. There were trade relations by sea, and relations of many sorts—largely hostile—in Turkestan.

With Western Turkestan

converted to Islam and under Arab rule, and Eastern Turkestan a part of the Chinese domain, there was naturally a constant inter¬ change—in the course of which papermaking entered the Islamic world. This intercourse was somewhat retarded by political condi¬ tions in Central Asia during the Sung period, only to be renewed and greatly increased under the Mongol empire. The extent of Arab penetration of China at this time is borne witness to by the fact that the province of Kansu, the main avenue of Arab trade, is still largely Moslem, and that the Moslems there, as they do all over China, have a large admixture of Arab blood. In fact, all large cities of China and many small ones have a consider-



[Ch. 15

able Moslem population, who trace their descent back to the inter¬ marriage of Arab traders with Chinese women during this period of Moslem penetration that reached its culmination during the Mon¬ gol dynasty. Commerce came by sea as well as by land, the coast cities of South China having been great Arab centers, and having today also a large Moslem population. This peaceful penetration of China by Arab trade is described both by Chinese and by Arabic writers, especially by those of Mongol times. Chao Ju-kua, who was a Chinese inspector of foreign com¬ merce in the province of Fukien some time during the half century before Marco Polo’s visit, has left a detailed description, too hazy for that of an eyewitness and evidently derived from Moslem traders, of the various lands of the West from Baghdad to Spain.4 Ibn Batuta, writing toward the end of the Mongol occupation, gives a wonder¬ ful picture of how all China was in his day permeated with Arabs. It is no longer a tale of marvelous things he tells. His description sounds as if such a trip as his were an everyday occurrence. In one city after another he is met by the Arab merchants and he notes that they are always organized under a judge and a Sheikh-ul-Islam. But most astonishing of all is the narrative where he tells of casually meeting a man at a feast in Hangchow and discovering that he and his new-found acquaintance came from neighboring cities in Mo¬ rocco, and that they had met a long time before in Delhi. The narrative ends, “I met his brother later in the Soudan; how far these brothers are separated, the one from the other.” 5 By the time of Ibn Batuta the world was already growing smaller, and consider¬ able information about China was part of the common knowledge of those who gathered about the bazaars of Tabriz and Cairo and Algiers. Yet in spite of all this intercourse with the Far East, books in Arabic were never printed. Whether, unrecorded and unheralded, there was an obscure block printing activity—the making of charms or playing cards—is another question and will be discussed later. But as far as literature is concerned, the Arabs did not print. Rashideddin, who was grand vizier of Persia during the Mongol period at just the time when Tabriz was the great bridge between the East

Ch. 15]



and the West, and who wrote a clear account of Chinese printing in his world history (see Chapter 17), seems never to have con¬ templated having his history printed. Instead, he provided in his will, and left funds for the purpose, that each year two full copies of all his works should be made by hand, one in Arabic and one in Persian, until gradually there should be a complete copy in the mosque of every large city of the Moslem world. Though Arab culture, which so profoundly influenced reawak¬ ened Europe, knew of Chinese printing, the refusal of its literary men to profit by the art made Islam on the whole a barrier rather than a bridge for the transmission of block printing to Europe. The story of the penetration of this barrier—by the Mongols from the East, by the Crusaders from the West—and of the obscure forms of printing that succeeded in spite of prejudice in finding lodgment in Moslem soil, will be told in the next chapters.

NOTES 1. It is probable that an edition of Rashid had already appeared at Con¬ stantinople in 1714 under similar auspices. These were the only two books published. 2. We are indebted to Dr. Grohmann for the statement that, according to Theseus Ambrosius (d. 1540) in the book Introductio in Chaldaicam Linguam (Pavia, 1539), folio ii recto, 84 recto, 200 verso, the father of Ales¬ sandro de Paganinis, who worked as printer in Venice between 1483 and 1499, printed the Koran with movable type. Unfortunately no copy of this first edition is extant. This was without doubt the oldest Arabic printing in Europe. At Fano, Arabic printing was done in 1514, and it seems probable that in 1518 another edition of the Koran was printed in Italy. 3. A. Vissiere, who contributed an essay on Chinese Moslem writings to d’Ollone, 1911:389-419, lists nineteen books printed by Moslems in China during the years in question and Pelliot, 1935:279, has added another, Chih sheng shih-lu (Biography of Mohammed), dating from 1775-85. In an edict of the Ch’ien-lung period, dated July 12, 1782, given in the Ta Ch’ing shih-lu, Kao-tsung reign 1158/23, is a reference to a Moslem from Yai-chou, Kuangtung, carrying twenty-one Arabic and several Chinese books, one of the latter being the Tien fang chih sheng shih-lu nien-p’u, which we take to be a



[Ch. 15

longer title for the same work. (This reference was furnished by Canon Claude L. Pickens, Jr., who has specialized in the Moslem impact on China.) Nothing is reported as to whether any of these books were printed. 4. For the English translation of Chao Ju-kua’s works, with valuable introduction and notes, see Hirth and Rockhill, 1911. 5. Yule, I9I3/i6:IV, 1-166. Ferrand, 1914:11, 432-33, has expressed doubts as to Ibn Batuta’s veracity, and Laufer, 1927:75, considers his account overvalued. Perhaps so, but it is worth noting that a Chinese scholar has come to Ibn Batuta’s defense. Chang Hsing-lang, 1938:611, shows, for ex¬ ample, that Shaikh Burhan-uddin of Zayton (see Yule, ibid., 120) is men¬ tioned by name as Hsieh pu-lu-han-ting in the Ch’uan-chou fu chih (edition of 1763) 75; further, the local history reports that he reached China in 131213, was asked to take charge of the mosque, and died in 1370.


knew almost as little of China as it did of Amer¬

ica, for Islam was a barrier well-nigh as formidable as the Atlantic Ocean. It was in the early part of the thirteenth century that Jenghis Khan and his Mongol hordes broke through this barrier, and Europe and China stood for a short time face to face. For a century or more —the middle of the thirteenth century to the middle of the four¬ teenth—the contact between Europe and Eastern Asia was far closer than ever before and probably closer than at any subsequent period down nearly to the nineteenth century. To travelers from the West, Cathay was the land of marvels, of wealth, and of intel¬ lectual culture—a land to be looked up to. For one century, and one century only, the way was wide open. With the fall of the Mongols the curtain fell, only to be raised a century and a half later, after Europe had passed through the Renaissance. In the year 1206 Jenghis received the submission of the Uigur kingdom and incorporated into his own rapidly expanding state the brains and the marvelously eclectic culture that had centered about the oasis of Turfan. One country after another was added to the ever-growing empire—parts of North China in 1215, Korea in the same year, Khwarezm (Russian Turkestan) in 1223, Persia in 1231, the balance of North China in 1234, Russia in 1240, Baghdad in 1258, and South China in 1279. Devastating raids were made into Poland, Hungary, Germany, Indochina, and Java and a great navy was sent against Japan. Almost the whole continent of Asia was under one rule, and with it was united much of European Russia. Roads were built and armies, mounted on fast horses, were con¬ tinually passing to and fro. In their wake came trade—overland T55



[Ch. 16

trade between the lands of the Near East and those of the Far East over the Turkestan passes and the Mongolian deserts flourishing as they have never flourished before or since. China and Europe met: a China that for three centuries had been printing books; a Europe that was just waking up to the need for books. Just at the end of the period of Mongol domination the first primitive block prints ap¬ peared in Europe. No clear documentary evidence can be produced to show how block printing entered, but certain phases of the history of the Mongol period show points at which Europe was especially exposed to Far Eastern influence. Upon these different phases are based various hypotheses as to route of transfer—hypotheses which are not mutually exclusive, and which may show a variety of influ¬ ences to which the beginnings of block printing in Europe were due.

As already explained (Chapter 14), it was through the Uigur Turks that the Mongols first came in contact with civilization and with the art of printing. One of the first tasks of Jenghis after he had received the submission of the Uigurs was the conquest of the kingdom or empire of Tangut, which had established itself for some two hundred years in northwestern China and eastern Mongolia. The Tanguts were a people of Tibetan stock, but the population over whom they ruled was largely Chinese and Tatar. Like the Uigurs the Tanguts were printers, using the art largely for the duplication of Buddhist sutras. Such sutras in a peculiar ideographic character have been found at Ning-hsia, Tun-huang, and Turfan, but a number too have come from Kharakhoto, far out in what is now Mongolia, where they were discovered by the Russian expedi¬ tion of Koslov. Here Buddhism was the religion of the state, and block-printed sutras, both in Chinese and in the Tangut language, were printed by imperial order. With these Chinese and Tangut siitras were found two sutras and some paper money in the language and the characters of the Mongols, showing how the conquerors took over the culture of the conquered.




As the Mongol hordes moved eastward they were constantly in touch with peoples who knew how to print, and as they adopted the culture of conquered lands, it was a culture based on printing that they adopted. As has already been pointed out in Chapter 10, printing in China had just reached its highest point of achievement at the time of the Mongol conquest, and during the period of Mon¬ gol control there was no diminution in the number of printed books produced. The Mongol rulers made it a point of honor to see that the ancient Chinese literature was printed not only in Chinese but in their own language as well.

After the conquest of North China, the Mongol armies turned westward, penetrating Persia and Russia and even Hungary and Poland. In the invasions of Hungary and Poland, Mongol domi¬ nation came nearest to the heart of Europe. The great campaign against Poland took place in 1241, immediately after the conquest of Russia.

Cracow and other leading cities were burned, Silesia

was invaded, and a combined German and Polish army was de¬ feated at Liegnitz in German Silesia. So great was the panic through¬ out Germany that the herring fisheries on the Frisian coast were abandoned and, according to a contemporary chronicler, herring about the coast of England became so plentiful that they sold in the English market for half their usual price. Meanwhile Hungary was invaded, Budapest burned, and the whole country ravaged, even a number of cities along the Adriatic coast being sacked. For¬ tunately for Europe, the death of the Grand Khan Ogatai recalled the Mongol armies. They occupied Hungary only a year, Poland a still shorter time. A second invasion of Poland took place in 1259, a second invasion of Hungary in 1285. In these invasions the capitals of the two countries were again burned, but in neither case was the occupation of long duration. In these campaigns the Mongol armies came very close to those places where the earliest block printing activities of Europe during the next century were carried on—Venice, Prague, and the cities



[Ch. 16

of Bavaria. Did they leave in their wake anything that suggested the art? The fact that the earlier and more important campaigns were fought before the Mongols had attained a high degree of civilization, and that all the campaigns were little more than raids without much opportunity for cultural mingling with the people of the land, suggests a negative reply. The communication of such objects as printed charms or playing cards is not impossible, how¬ ever, and more important printed matter, such as religious pictures, may have been in the hands of Uigurs who accompanied the Mon¬ gol armies.

The influence of the Mongol occupation of Russia was far dif¬ ferent. Russia was invaded in 1223, conquered in the campaigns of 1236-40, and held in Mongol hands for more than two hundred years. While Russia was never as directly controlled as were China and Central Asia and Russian princes had considerable autonomy, circumstances made necessary a large amount of travel between Moscow and the court of the Grand Khan. Every Russian nobleman of the higher ranks was compelled to go to Karakorum for inves¬ titure, at least during the early part of the occupation, and many internal disputes had to be referred to the Grand Khan for decision. Throughout Mongol times the market of Nizhni Novgorod, east of Moscow, was a distributing center for articles from the Far East entering Europe, and here the caravans of China and Turkestan came in contact with the river-borne traffic of the cities of the Hanseatic league.

One section of Novgorod is still called the

“Cathay Section,” and an important street of Moscow “Cathay Town” (i.e., Kitay Gorod). While the hypothesis of Russian agency rests in the main on circumstantial evidence provided by the general history of the period, there are in addition certain clues, a further investigation of which may lead toward more direct evidence. The seal cutter of the Grand Khan Kouyouk (1246-51) is known to have been a Russian by the name of Cosmas,1 a fact of importance

Ch. 16]



in consideration of the close connection that existed between seal cutting and block printing. Furthermore, de Rubruquis states that the currency of Russia under the Mongols consisted of bits of leather or fur “marked with colors.” 2 Whether by this is indicated a stamping process or any¬ thing allied to printing—after the analogy of the printing of paper money that was going on in the other parts of the empire—is un¬ certain, but in any case the paper money of China and Central Asia, parts of the same empire and closely connected by trade routes, could not have been unknown in Russia. The statement that printing came into Europe from China by way of Russia is first made by the historian Jovius in 1546, just a century after Gutenberg, in what is apparently the earliest reference to Chinese printing in European literature. Jovius’ statement is: There are there [at Canton] printers3 who print according to our own method books containing histories and rites. . . . Pope Leo has very graciously showed me a volume of this sort, given as a pres¬ ent with an elephant by the king of Portugal. So that from this we can easily believe that examples of this kind, before the Portuguese had reached India, came to us through the Scythians and Muscovites as an incomparable aid to letters.4

John of Plano Carpini was sent by Pope Innocent IV in March, 1245, on an embassy to the court of the Grand Khan. He went by Prague and Kiev to Mongolia, where he presented his letter and re¬ ceived his reply. This reply—the original—was discovered by acci¬ dent in the year 1920 in the archives of the Vatican.6 It is written in Uigur and Persian and contains in lieu of his signature the seal of the Grand Khan Kouyouk (grandson of Jenghis). This is the first recorded appearance in Europe of an impression from a seal based on those in use in China and impressed with ink upon paper.6 The seal was without doubt made by Cosmas, the Russian seal cut¬ ter, of whom Plano Carpini tells. This letter, written in the Persian and Uigur languages, sealed with a Mongol seal of Chinese style



[Ch. 16

that had been cut by a Russian seal cutter, and sent by the hand of an Italian monk to the Pope, is a typical example of the cosmopolitan character of the Mongol Empire, bridging the gap between the East and the West. In 1248 and 1253 two embassies were sent by Saint Louis of France, then in Cyprus on a crusade, to the court of the Grand Khan. The leader of the second of these embassies, William de Rubruquis, in his description of the journey tells of the number of Europeans whom he met at the Mongol capital. Among the pris¬ oners who had been brought from Belgrade and from Hungary and who were still living at Karakorum were the nephew of the Nor¬ man bishop of Belleville near Rouen; a French woman from Metz named Paquette who was married to a Russian; an Englishman named Basile; and a Paris jeweler, Guillaume Boucher, who was serving as goldsmith to the Khan. Other Westerners at Karakorum in the narrative of de Rubruquis were a Christian from Damascus and an Armenian bishop. The knight Baldwin of Constantinople had just left the court with another Knight Templar.7 All this indi¬ cates that even at the beginning of the Mongol regime the men who wrote books were not the only people who went back and forth be¬ tween the Mongol court and Europe. De Rubruquis, while not describing printing, is the first Euro¬ pean writer to mention printed paper money. In the same section in which he mentions the leather money of Russia, he says, “The ordinary money of Cathay is made of cotton paper,8 as large as a hand, upon which they imprint certain lines like the seal of Mangu (imprimunt lineas sicut sigillum Mangu).”9

Marco Polo was the one traveler in Central Asia and China who wrote such a clear account of his travels as to make a deep impres¬ sion on Europe. For this reason a great variety of things that have come from China to Europe have been credited to him, and block printing is no exception. The story is that a certain Pamfilio Castaldi of Feltre, a block printer at the end of the fourteenth century, had




learned the art from seeing some pieces of wood that Marco Polo brought back to Venice and that had served for the printing of Chinese books. The story, while not inherently impossible, rests on insufficient foundation.10 It is a strange fact that Marco Polo’s de¬ tailed description of China never mentions printing, except in the passage already quoted in Chapter n on paper money, and there his interest is not in the printing but in the money. If the tradition mentioned above is in any way founded on fact, it is more likely that the blocks seen by Castaldi were brought from China by one of the many nameless travelers who came back to Italy from the Khan’s dominions during the half century or more after Marco Polo’s return, rather than by Marco himself.

The men of education in mediaeval Europe—the men interested in books—were primarily priests and monks. If the bulk of all scien¬ tific study of the life, customs, and history of China in later times up to the beginning of the nineteenth century was done by Roman Catholic missionaries, the same must have been still more true in a day when the laity were largely uneducated. The first missionary sent by the Pope to China, John of Monte Corvino, arrived in Cambaluc about 1294, just after Marco Polo left for Europe. He remained at Cambaluc as head of the mission till his death in 1328. In 1305 he wrote home that he had already baptized six thousand converts, that he had built a church in Cam¬ baluc, that he had learned the Tatar language and had translated into this language the New Testament and the Psalter. The next year he wrote that he had built another church in Cambaluc on land presented by a resident Italian merchant, and that he had prepared six pictures, representing scenes from the Old and New Testaments, for the instruction of the ignorant, with explanation in Latin, Tursic,11 and Persian letters. In 1307 Pope Clement V raised John of Monte Corvino to the rank of archbishop, and sent seven Franciscans with the rank of bishop to assist him. Only three arrived; they worked for five years



[Ch. 16

in Peking, living on a subsidy from the Khan, then moved to Fukien, where a strong mission was established and a church built with funds given by a local Armenian woman. There were missionaries of the Roman Church at the same time at Yang-chou and in Tur¬ kestan.12 These missionaries, spending their lives in China, learning the language and mingling with the people, must have come in contact with printed literature at every turn. John of Monte Corvino in the first dozen years of his work, even before reinforcements had arrived, had already translated the New Testament and Psalter, and prepared pictures and text for the ignorant at just the time when in China it was the natural thing to have every important literary work printed. There is no question that the Chinese who were associated in the work of translation would have suggested that the translations and the pictures should be brought before the public in what to them was the usual and natural way. Whether the missionaries agreed and thus became the first European patrons of the art of printing, we have no means of knowing. That religious image prints, pre¬ pared, like the pictures of John of Monte Corvino, “for the ignorant,” began to appear in Europe some time within the half century after these early missionaries laid down their work, may not be alto¬ gether a coincidence.13

As has already been mentioned, a Russian seal cutter, a Paris goldsmith, and a number of other Europeans were already in the middle of the thirteenth century at the court of the Grand Khan in Mongolia. Marco Polo tells of a German who assisted Kublai’s gen¬ erals in the preparation of engines of war. But it was during the first half of the fourteenth century, after Marco Polo’s reports of Cathay’s wealth, that trade between Europe and China multiplied.14 The extent of that trade can best be understood by a study of the zeal with which Columbus and his successors more than a century later were ready to brave untold hardships to rediscover the wealth of the “Indies” and find the Northwest Passage to Cathay. The




traders of Mongol times were not men of letters and there are only a few data, largely furnished by missionaries, from which to form a picture of this early commerce. Andrew, bishop of Zayton in Fukien, wrote in 1326, quoting the opinion of Genoese merchants at that port about exchanges. Odoric, missionary in China from 1323 to 1327, referred for confirmation of the wonders he related about Kinsay (Hangchow) to the many persons whom he had met at Venice since his return who had themselves been witnesses of these marvels. Marignolli, writing after his return from China in 1346, told of the fondaco or “factories” he found attached to the convents at Zayton for the accommodation of Christian merchants. But per¬ haps the best indication of the extent of European trade with China at this time is contained in a handbook prepared by Pegolotti in Florence in 1340.15 This book, which is a trade guide to the various ports of the world, devotes its first two chapters to Cathay, giving such information as a European merchant traveling in that country would need to know—about routes of travel, about imports and ex¬ ports, about currency, weights, and measures, taxes and duties, etc. Like Marco Polo, the writer of this book describes Chinese paper money, even giving particulars about rates of exchange, but—like Marco Polo again—what interests the writer is not the printing but the value of the paper money. There is no record to show that printing was brought from China to the West in the wake of trade, nor is it likely that merchants would have come as closely in contact with Chinese printing as would missionaries and translators. Yet the very fact that, during the half century before block printing appeared in Europe, large numbers of obscure men whose names have not been recorded in history were moving back and forth between China and Europe both by land and by sea, is not without significance. In a later period, when the way to Cathay had been rediscovered by Vasco da Gama, and trade had been re-established—some half century or more after Gutenberg—a Chinese printed book found its way very quickly to Portugal and was presented by the king of Portugal to the Pope. It is not an unlikely hypothesis that a specimen of Chi¬ nese printing or a report of Chinese printing, brought to Europe



[Ch. 16

during the earlier period when trade was more extensive, was one of the influences back of the great block printing activity that pre¬ ceded the invention of type.

The one point at which Europe and Eastern Asia came together and mingled most fully was Persia. The significance in the history of printing of the interchange of ideas between East and West that took place in Persia, and especially in the great cosmopolitan center of Tabriz during the enlightened reign of Ghazan Khan and under other Mongol rulers, is so great that a special chapter must be devoted to this natural crossroads between the East and the West.

NOTES 1. For this statement of John of Plano Carpini, and for discussion of its implications, see Pelliot, 1922-23:27-28. An impression from this same seal, found in the archives of the Vatican, is reproduced (with translation) in the same article. 2. The currency of Russia in earlier times had been the furs of animals, especially the Siberian squirrel, which were worth an exact weight of silver. It is generally believed by Russian writers that during Mongol times, under the influence of the paper money of the rest of the Mongol empire, the Rus¬ sians began to use, instead of whole furs, small pieces of fur stamped by the government and redeemable in the stores of the government for whole skins. The matter is the subject of some debate. It is discussed with a full bibliog¬ raphy in von Ebengreuth, 1904:36. 3. The Latin is typographos artifices. Bernard, 1945:2, thinks that this means “artisans de typographic” and not “xylographie.” 4. “Quod maxime mirandum videtur, ibi [Canton] esse typographos artifices, qui libros historias et sacrorum ceremonias continentes, more nostro imprimant: quorum longissima folia introrsus quadrata serie complicentur. Cuius generis volumen a rege Lusitaniae cum elephante dono missum Leo pontifex humaniter nobis ostendit: ut hinc facile credamus eius artis exempla antequam Lusitani in Indiam penetrarint per Scythas et Moscos ad incomparabile litterarum praesidium ad nos pervenisse.”

Paulus Jovius (Paulo

Ch. 16]



Giovio), 1558:161. This earliest European mention of Chinese printing has apparently not before been noticed except in an unpublished manuscript in St. Bride’s Library, London, by Richard Smith, written in 1670, in which Jovius’ view that printing was introduced from the “Indians of Cataia” by means of “the Scythians and Muscovites” is rather unfavorably discussed. Jovius had been an ambassador to Moscow not long after the new Russian state had freed itself from Mongol domination, and has left a history of Russia as well as several books descriptive of that country. His statements concerning Russia therefore carry considerable weight. On the other hand, he quotes no authority and his statement may be only a conjecture based on his general knowledge of Russian history and of Chinese printing. Note also that in 1562 Busbecq wrote a letter regarding the Chinese art of printing; see Sarton, 1942:562. Montaigne likewise was struck with aston¬ ishment over the fact that the Chinese for a thousand years (“mille ans auparavant”) had enjoyed the printed page; see his Essays of 1588, cited by Bernard, 1945:3. 5. This letter from the Grand Khan to the Pope was discovered in the Archivio di Castello by P. Cyrille Karalevskyj. It was identified and de¬ ciphered by Pelliot, and has been published by him, together with a facsimile of its seals. (Pelliot, 1922-23:3-30.) The seal impressions, like the Chinese seal impressions on the letters from the Persian Ilkhans (see Chapter 17, note 2) are 5% inches square, but these are in Mongol, not Chinese. Pelliot (pp. 27-28) has given his reasons for concluding that these seal impressions were made from the seal described by Carpini and cut by the Russian seal cutter Cosmas. Some half dozen other letters from Mongol sovereigns (most of them from Ilkhans of Persia) have also been found in the Vatican archives, and are published in subsequent numbers of the same review. 6. A number of Chinese seals were dug up in Ireland about 1800 and are described in a paper read before the Belfast Literary Society by Edmund Getty in 1850, entitled, “Notices of Chinese Seals found in Ireland.” Getty believes that they were brought to Ireland by early monks and date from the eighth or ninth century, but it is more than probable that they were brought by Irish sailors at a much later date. 7. For a recent discussion of the Europeans at Karakorum, see Olschki, 1946. 8. Here de Rubruquis seems to have erred. Pelliot, 1922-23:14, reports no cotton in the letter of 1246 to the Pope. Berthold Laufer points out that the paper money of the Yuan and Ming was made of mulberry pulp. (See Cordier, 1920^70-72.) Vidal and Bouvier, 1925:161, also find no cotton in the paper of 1600, apparently made by the Chinese, which was used as a specially luxurious writing paper by the Persians of the sixteenth century and later. 9. De Rubruquis, Latin edition of d’Avezac, 1839:1V, 329. Additional interest attaches to the reports of De Rubruquis, on account of the fact that



[Ch. 16

Roger Bacon read his book and was personally acquainted with him after his return from Central Asia. (See Bridges, 1897:1, 353_66.) 10. It is Pauthier’s edition of Marco Polo (G. Pauthier, Le livre de Marco Polo, p. 78) that has given currency to this story. Pauthier’s statement is quoted from Octave Delpierre, Analyse des travaux de la Societe Philobiblon a Londres, p. 23, which is in turn quoted from Curzon, 1860:25. On what Curzon based his statement is uncertain. It seems probable that it was an old Italian tradition. Yule, 1903:1, 138-41, scotches it but is quite ready to believe that Christian missionaries or traders from Europe may have brought home “block-books” from China during the fourteenth century. 11. The signification of “Tursic” is uncertain. Moule, 1930:178, suggests that it appears to represent a Persian word applied contemptuously to Christians and people of other religions; hence, Mongol or “language of the idolators. 12. The later history of this first phase of Catholic missions in China is shrouded in mystery. On the death of Monte Corvino in 1328, Friar Nicholas of Paris was sent out from Avignon to succeed him, accompanied by twenty monks and six lay brothers. They left Avignon in 1333, and in 1338 are heard of at Almaligh in Eastern Turkestan. By this time Islam was rapidly gaining ground in Eastern Turkestan and the land route was becoming increasingly difficult. There is no record that they ever reached China. However, in 1338 Europeans arrived in China with letters written in 1336, and again in 1342 the Pope sent an embassy, headed by Marignolli, who after four years returned to Europe and wrote an account of his journey. After Marignolli’s return in 1346 nothing further is known with certainty of the mission in China, though there are indications that the last missionaries in Fukien were martyred in 1362. From the Avignon end it is known that more missionaries were sent out. William of Prato was made archbishop of Cambaluc in 1370 and sixty clergy followed him. Francis of Podio was sent the next year as apostolic legate with twelve followers. The Vatican records show a full line of arch¬ bishops of Cambaluc through the next century. But, so far as is known, they went out into the darkness, never to be heard of again. The breakup of the power of the Ilkhans of Persia and the renewed activity of the Turks closed both the land route and the water route between Europe and the Far East, while the fall of Mongol power in China in 1368 rendered China in¬ hospitable to foreigners. For a century and a half the barrier between China and the West was seldom crossed. Columbus tried to reopen a route for inter¬ course in 1492; Vasco da Gama succeeded in 1499. But even after the dis¬ covery of this lengthy route around Africa, it was centuries before China and Europe came again so close together as they had been during the time of the Mongols. 13. Further exploration of libraries and archives in Italy may add evidence with regard to this hypothesis. In 1922, in the Laurentian Library in Florence there was discovered a Latin manuscript Bible that had been in use in China by missionaries of the Mongol period. Unfortunately none of the Chinese or




Mongol Christian literature that they prepared has yet been discovered. It would also be interesting to learn how many of this company of missionaries returned to Europe when the Mongol Empire broke up, and what they did after their return. 14. In 13055 John of Monte Corvino had seen no European for twelve years, though a “Master Peter,” a “great merchant,” had accompanied him to Cambaluc. It was probably between 1310 and 1320 that commercial inter¬ course on a larger scale began. x5- Pegolotti’s famous book has received a new rendering in Evans (ed.),



From the

days of Mohammed until the time of the Mongol con¬

quests, the world of Europe and Asia was divided into three very distinct cultural areas—Christendom in the West, Islam in the center, and the Buddhist and Confucian domain in the East. In Persia during the Mongol regime the three for a time seemed almost to coalesce. Under the tolerant rule of the Ilkhans, Buddhist and Mos¬ lem, Christian and Jew succeeded each other in the highest positions of the state with surprising swiftness, while all races of the known world mingled in Tabriz, the cosmopolitan capital. Persia was first overrun by Jenghis in 1221, and in 1231 was brought fully under Mongol domination. In 1258, Baghdad was taken by the great Mongol general, Hulagu, brother of Kublai, and Mesopotamia with much of Syria and Armenia was added to the Mongol domain. This brought the Mongol armies face to face with the Crusaders. Certain of the Mongol allies even proceeded as far as Palestine and sacked Bethlehem, the Crusaders’ chief shrine. But as a rule the Mongol Ilkhans (as Hulagu and his successors were called) were more or less allied with the Crusaders against their common enemy, the Saracens. Constant embassies were exchanged between Tabriz, the Mongol capital of Persia, and the later Crusad¬ ing princes. In the letters that have been preserved, the Mongols with true diplomatic courtesy express their deep attachment to the Christian faith, and the replies of the Crusaders greet them as Christian brothers, as do also letters from James of Aragon and Ed¬ ward II of England. A number of embassies were even sent to Europe by the Mongol rulers of Persia, bearing letters to the Pope, to the king of France, and to the king of England, and several such letters with their large vermilion seal impressions in Chinese char168

Co • cto 03 ^g E» •S J ■£ 8 < ^ Sdi - -Q }J

cr3 pL, ^ Eo . . GOJ *-M & g ^ o

to .GG _ o «-* P H3o X

5 ^

ww X £ £ o o z



> X o

>w PS

P C —

O ✓—s

3 §

G ro *V 3 X H m

aj M-(

c §

3 ? o

3 9 -Q

5< n ° *• o xo

s^ „2o *-' X< a Ph ^ cow Xip w


o aX O

Wooden Types


Impressions of the Early Fourteenth Century.

These types, found

by Professor Pelliot in one of the Caves of the Thousand Buddhas, date from about the year 1300. They are in the language of the Uigur Turks. Each type represents a word.


to paper 2.2 cm.; width 1.3 cm.; length varying from 1.0 cm. to 2.6 cm., according to length of word.)

East Asiatic Library, Columbia University.

Early Metal Types from Korea.

Ch. 19]



and detailed references to card playing begin, references which can¬ not possibly be confused with chess or dice, and which show a widespread and highly developed game of cards that continued throughout the Southern Sung and Mongol periods. Our next sources for the history of playing cards are European. In all mediaeval Arabic literature they are, so far as is known, never mentioned.1- Nor is archaeology of very much aid, for though two Chinese playing cards—presumably ancient—were turned up by the German expedition near Turf an, there is no material by which they can be dated with certainty. For further clues it is necessary to turn to Europe. Here again in the early records there is confusion of terms. Chess, which had spread from India through the Saracen world, had reached Europe with the first Crusades13 or even earlier, and had been played for two or three centuries before cards began to appear. The Persian origin of the European game of chess is clearly indicated in the use of such words as check’ {shah, king) and ‘mate’ {mand, at a loss).14 Certain supposed early references to cards in Europe, during the thirteenth century and the early part of the fourteenth, are now generally conceded to refer to chess rather than to cards.15 The earliest references to playing cards in Europe that can be clearly differentiated from chess, follow each other with rapid suc¬ cession in various countries—Germany 1377,16 Spain 1377,17 Luxem¬ burg 1379,18 Italy 1379,19 Belgium 1379, France 1382.20

By 1397

card playing had become so popular in Paris as to occasion an edict by the provost of the city, in which workingmen were forbidden to play cards and certain other games on working days. In 1404 the Synod of Langres forbade the clergy to play cards. A climax was reached in 1423, when a famous sermon against card playing was preached by St. Bernard of Siena from the steps of St. Peter’s in Rome, with the result that his hearers rushed to their houses, brought such cards as they possessed to the public square and burned them. A comparison of the dates of the spread of playing cards with the dates of the earliest religious block printing is significant. The most generally received view is that the earliest religious block



[Ch. 19

prints date from the last decades of the fourteenth century. A more advanced stage of the art is shown in the “Virgin and Child of 1418, now in the Brussels Royal Library, which preceded St. Bernard’s sermon against card playing by five years. The period during which playing cards spread through Europe corresponded therefore with the period of the earliest religious prints. It corresponded also with the half century after the collapse of the Mongol Empire. The question how early the playing cards of Europe began to be printed has been much debated.21 The consensus is that very early in the fifteenth century or even before 1400, and possibly from the time of their first use in Europe, at least some of the cards were printed. The printing of cards soon came to be an important indus¬ try. An edict of the Council of Venice, dated 1441, indicates that the card printing industry, which before that time had flourished in Venice, was already being interfered with by outside competition." Card makers, who were presumably card printers, had already be¬ tween 1418 and 1438 been mentioned five times in the city records of Augsburg and Nuremberg, and at about the same time the records of the city of Ulm in Germany show that cards were being shipped in barrels to Sicily and Italy. By some the first printing of playing cards is believed to have preceded the making of image prints, but it seems more probable that the two forms of printing developed side by side at about the same time, and that they were sometimes carried on by the same persons.2 ' In determining the source of European playing cards, the stu¬ dent is faced with a paradox. With the exception of a seventeenth century writer who claims that cards were introduced directly from China to Italy,24 European sources are unanimous in indicating that the game was derived from the “Saracens.” On the other hand Arabic sources, so far as has yet been discovered, never mention the game. From them it would seem as if the injunctions in the Koran against games of chance were obeyed to the letter. To solve this apparent paradox, it is necessary to understand the historical situa¬ tion. Card playing had been general in China for at least two cen¬ turies before it was known in Europe. Presumably cards were in common use in the Mongol armies and among their camp followers.




Chinese, Central Asiatics of various sorts, Moslems, Genoese, and Venetians were living and trading together for a century or more in Persia.

Immediately thereafter cards spread through Europe

from the

land of the Saracens.” But card playing had not, like

chess, made itself sufficiently at home in the Islamic world to enter into Arabic literature. The game passed quickly and lightly across the Near East without leaving any trace in Arabic records. There was no repetition of paper’s laborious course of a thousand years from China to Europe, for the things that Europe had real demand f°r

such things as gunpowder and playing cards—it got quickly!

Furthermore, the Mongol conquests and the Crusades had inter¬ vened. Of the connection between the introduction of playing cards from China through the Islamic world to Europe and the trans¬ mission of block printing there is no certain proof. The cards of China were of course printed. It is difficult to imagine the colonies of Chinese and Central Asiatic merchants who settled in Tabriz not bringing their games and their gaming habits with them, as Chinese have always done in those countries to which they have gone in more modern times. It is easy to speculate further, to think of Chinese block cutters (perhaps the same men who made the illfated paper money of Tabriz) setting to work to make cards and thus to avoid the long and expensive import across the deserts, and gradually adapting those cards more and more to the language and customs of the land in which they lived. It is certainly not difficult to imagine the joy with which Venetian merchants and perhaps some of the later Crusaders hailed this new-found game and brought it back with them to Europe. And it is not likely that the cheap and simple way in which the cards were produced would have escaped them. But all this is conjecture. What is known with certainty is that printed playing cards were in common use in China before the Mongol conquest; that immediately after the Mongol period cards began to appear in Europe and were recognized as of Eastern origin; that, either from the first or soon after, these cards were printed; that playing cards were among the first, if not altogether the first,



[Ch. 19

block prints in Europe; and that the printing of cards constituted an important industry both in Venice and in southern Germany in the early part of the fifteenth century. While it is not safe to say with certainty that playing cards in coming from China to Europe brought block printing with them, the evidence is at least sufficient to suggest that among the possible ways by which block printing may have entered the European world, the use of playing cards holds an important place.

NOTES 1. Laufer, 1927:76, holds that backgammon was introduced to China early in the sixth century. Yoshita Harada, more precisely indicates a game, introduced from a foreign country, called wu shuo, mentioned in the art section of the Wei shu (compiled by Wei Shou, 506-72), which he believes to be ancestral to backgammon. The general name for the game under the T’ang was shuang lu which is not to be confused with liu po, a much earlier game of indigenous origin. See L. S. Yang, 1947:202-6; and 19523:124-39. 2. From very early times (at least from the middle of the Chou dynasty) there had been a game of ch’i in China which is commonly translated chess. The history of this game is discussed by Parker, 1889:54. This game has sur¬ vived under the name of wei ch’i. It is much more complicated than our chess. The history of the Indian game of hsiang ch’i or “elephant chess,” which is more analogous to our own game, has been traced in Chinese and Manchu sources by Himly, 1897:155-80; and 1887:461 ff. According to Himly, the first reference to the Indian game in Chinese sources is in the year 569, and is quoted in the T’ai p’ing yii lan Encyclopedia of 983. Tu Ya-ch’iian, 1933:11, depending on another source, puts the date a year earlier. H. }. R. Murray, 1914; 123, holds with Himly that the first certain account of chess occurs in the Yu \uai lu or Book of Marvels, written by Niu Seng-ju (778-847). It seems to have migrated from India both to China and to Persia in the sixth and seventh centuries, and to have reached Japan also before the end of the seventh century. The above conclusions must be taken tentatively. Confusion of terms between the Chinese and Indian games makes it difficult to trace the early history of chess in China with certainty. Parker finds a certain amount of indication that hsiang ch’i was not an importation from India but had an independent Chinese origin, going back possibly even farther than wei ch’i. In any case, it was in all probability the Indian game rather than the Chinese that was the ancestor of the Arabic game, which, in turn, was the ancestor of European chess.

Ch. 19]



Some points in the Chinese game as played today are of historic interest. The castle, called chii, chariot, is the most powerful piece, as with certain limitations, it can jump over intervening pieces like the knight. The knight’s move is the same as ours, and the piece is called ma, horse. There is an addi¬ tional piece called p ao, a word which in modern Chinese is written with a nre radical and means “cannon,” but which before the invention of gunpowder was written with a stone radical and meant a mechanical device for hurling stones. The transitional form of warfare finds an echo in chess, for the blue p ao is written with a stone radical, while the corresponding red piece has the fire radical. Likewise, the bishop, which has the same move as ours, retains on the blue side of the board its Indian name of “elephant,” while on the red side it changes to a word pronounced exactly the same but meaning “prime minister. As in the Indian, Arabic, and early European games, there is no queen. 3. Laufer, 1932:13-14, 43-45; and Goodrich, 1938:27, 38-40. The Chinese did not use the term polo (or anything like it) for their game; but the Tibetans, from whom the Chinese learned the game, did. It related to the ball, and in a wider sense to the game itself and to the polo ground. See also Laufer, 1916: 540-42. 4. Yoshita Harada, 1939:65, writes that some people credit Ts’ao Chih of the third century a.d. with the invention of dice, “but the guess is very doubt¬ ful. Dubs, 1938:1, 292, draws attention to an even earlier game which may have involved dice, and Waley, 1933:140, suggests that the 1 (or Book of Changes) indicates “an amalgamation of four different divining techniques,” one of which may have been derived from dice and dice throwing. Laufer, 1927:75, says that the game is mentioned in a.d. 406 in the Chinese version of Brahmajala-sutra (section 33), translated by Kumarajlva (d. 413?). The Chinese term is a transcription of Sanskrit prasa\a or paqa\a (die, dice). 5. “In the annals of the Ch’i dynasty under the date 501 a.d. it is stated: ‘According to T’ao Shih-hsing dice are a foreign game, which Lao-tzu found when he was in the land of the Western Barbarians. In recent years officials are playing it very much. How does it come about that they waste their time with foreign things and do not help their own country?’” T’u shu chi ch’eng, 807/6. Laufer, 1927:76, remarks that there is no doubt that dice in China are of Indian origin. “In India, they are of immemorial antiquity, being used both for divination and gambling.” 6. Hummel, 1939:265, writes that “nearly all manuals on dominoes, of which there are many, point back at least to the year 1120 a.d.” 7. Various origins of the names, yeh tzu hsi and yeh tzu {o, “leaf-game,” "leaf dice,” or “sheet-dice,” have been suggested. According to Ou-yang Hsiu (1007-72), Kuei t’len lu 2/13^ “The books of the T’ang dynasty were all in the form of roils. When it was necessary to have any written matter ready for quick examination, it was made on pages. In the same way, in order to have dice in a convenient form, they were made on cards, and this was the origin of the



[Ch. 19

word, yeh tzu hKo (from yeh tzu, a leaf or page). Before the end of the T ang dynasty there were already such ‘leaf-dice. From other sources it is known that the transition from rolls to paged books was due to the influence of print¬ ing-

. , The conclusions here stated with regard to the origin of playing cards, dominoes, and mah-jongg are based on somewhat obscure sources, and, while they are believed to be correct, they are stated with considerable reserve. It may be that the influence of paper money on the origin of playing cards should be more stressed. For an able exposition of this view see Culm, 19248. Unfortunately nothing is said here about printed “sheet-dice. Twice in the Kuei t’ien lu passage translated above the word hsieh, meaning to

write,” is used. 9. The correct pronunciation is ma-chiang, also known as ma-ch iao. 10. For Remusat’s statement and for translation of the passage in the Cheng tzu t'ung Encyclopedia from which it is taken, see Wilkinson, 1895:6178. 11. The Liao shih 7/a sentence runs as follows: “On the day chia-wu of the first month of the nineteenth year of Ying-li the emperor played yeh-ho with his courtiers.” Cf. entry under a.d. 969 in Wittfogel and Feng, 1949:257. Laufer, 1927:76, holds that the game was fully developed in the course of the ninth century, and refers to Schlegel, 1869:20. 12. This is the conclusion reached on the subject by such eminent Arabists as Dr. Grohmann of Prague and Cairo, Dr. Margoliouth of Oxford, and Dr. Moritz of Berlin. Laufer, 1927:76, notes that the Arabs transmitted playing cards to Europe, for the Spanish-Portuguese word naipe and Old Italian naibi are derived from the Arabic la ib or “play.” Blum, 1940:43, modifies this statement slightly maintaining that the Spanish word naypes comes from the Arabic naib. In all likelihood the game is not mentioned by the Arabs because it was forbidden in Islam. 13. It is probable that the first mention of chess in Europe is in the year 1061, and thus antedates by a few years the first Crusade. The Crusades and the Christian conquest of Spain had the effect of spreading the game through Europe. 14. The very early introduction of chess into England is indicated by great variety of uses of the word “check” (including “bank check” and verb “to check”) all of which go back to chess and so ultimately back to Persian word shah. The popular etymology, mate from mat (dead), is longer held. Cf. Maghadam, 1938:662-64.

the the the no

15. Here are included the supposed references to cards in England in 1240 and 1278, in Germany in 1291 and 1300, in France in 1328 and 1376. 16. British Museum MS. Egerton 2,419.

Ch- J9]



17. Prohibition of cards by John I of Castile. 18. Account book of Jeanne, Duchess of Brabant, May 14, 1379. 19- In the year 1379 was brought into Viterbo the game of cards, which comes from the country of the Saracens, and is with them called noth.” Covelluzo of Viterbo, writing in the fifteenth century, on the authority of a chronicle of one of his ancestors. 20. For dates of earliest references in Belgium and France see Blum, 194043-49-

21. The earliest references to playing cards give as a rule no clear indication of the method of manufacture. Even the order given in 1392 for three packs of cards for the King of France, which uses specifically the word painter, gives no suggestion what kind of cards were being used by the common people, for painted cards were used by royalty long after printing began. The earliest cards extant some printed, some painted, some printed in outline and filled in with a stencil also shed little light on the question, for they cannot be dated. They indicate merely that cards were being made in several different ways, pre¬ sumably at the same time, according to quality and price. Printing in its beginnings—whether of pictures, texts, or cards—was always the poor man’s friend. Laufer, 1927, is not convinced that playing cards were instrumental in transmitting the Chinese method of block printing to Europe. Blum, 1940:48, reports that the first cards were not printed as woodcuts before the middle of the fifteenth century. 22. In 1441 the Council of Venice issued the following decree: “Whereas, the art and mystery of making cards and printed figures, which is in use at Venice, has fallen to decay, and this in consequence of the great quantity of printed playing cards and colored figures which are made out of Venice, to which evil it is necessary to apply some remedy, in order that the said artists, who are a great many in family, may find encouragement rather than foreigners: Let it be ordained and established, according to the petition that the said masters have supplicated, that from this time in future, no work of the said art that is printed or painted on cloth or paper—that is to say, altar-pieces, or images, or playing cards, or any other thing that may be made by the said art, either by painting or by printing—shall be allowed to be brought or imported into this city.” 23. From the decree of 1441 several things are evident, first, that at some time before 1441 the printing of playing cards had been a thriving industry in Venice; second, that both printed and painted cards coming from some other place had interfered with that industry; and, third, that the printing of playing cards and the printing of saint images were closely connected. A source of these imported cards is indicated by an entry in the Red Book of Ulm in southern Germany, according to which, at just about this time playing cards were being shipped in barrels to Sicily and Italy. The city records of Augsburg



[Ch. 19

and Nuremberg mention card makers in 1418, 1420, 1433* 1435> an




O .ti



**3 > 3 X ,









8 w


G On

Nh T3





o G






s (/5



On O


o s



Q < ^ Q


!2 T3 « -deg



H v2







oo VO CO

§ | g -s G


rt 0-^2 o w «


« jp

• oo On cs


p W

.3 -2 -G U

rs 'u,





00 *P



a u



V T3 G G rt rt





a b C G




o w


IA C P O O ’£ CO CO t-H OO aj - qj





S ■§ rt

O -rs bO oj



3 e

g ^

bo C .i

B •«





P "P



3 *43

o rt Mrt 0 O -G .G q-> 0 G z,

s s ^



■a -o G


* U

c -3

V V G O >% H G

G ^ C G G >,

-G G

QJ bo


a, £ 3 >~. qj



o jp

t>c -a rt *3

"O -0

* eg



ro _C


O. .5

C os

lx a.

° ^ .5

SO c ^ O,


s 2 &

C -g

W -Q



5 ‘w O c o;


^ to

3 .5

03 w