People and Culture

Shoju Ikeda

Deciphering Old Manuscripts and Sharing Their Electronic Texts

Shoju Ikeda , Professor

Graduate School of Letters (Department of Humanities and Human Sciences, Faculty of Letters)

High school : Tochigi Karasuyama High School

Academic background : Graduate School of Letters, Hokkaido University

Research areas
Japanese Linguistics
Research keywords
Linguistics, data base, Kanji, dictionaries, electronic texts

What are you aiming for?

I aim to decipher books written over several hundred years ago in Chinese and Japanese characters that are different from those used today, create electronic texts after the original text has been deciphered, and build a comprehensive database consisting of these electronic texts which is open to the public on the Internet for international sharing.

The books to be deciphered are early “dictionaries” that nobody has succeeded in deciphering in their entirety.


What are early “dictionaries”?

Figure 1: Waeigorin Shusei, Second Edition (1872)
(Meiji Gakuin University Library digital archives )

Figure 2: Shinsenjikyo (housed in the archives
and mausolea department of
the Imperial Household Agency)

Thinking about the word “dictionary” as a Japanese, you may think of English-Japanese dictionaries, Japanese-language dictionaries, and Chinese-Japanese character dictionaries. Recently people generally use an electronic dictionary that combines multiple dictionaries and free dictionaries are available on the Internet. However, a dictionary used to be compiled into book form in the past.

For example, an English-Japanese dictionary that was used by a student of Sapporo Agriculture College, the predecessor of Hokkaido University, is held as rare book in the University Library (Sapporo Agriculture College Bunko). The dictionary is Hepburn’s Waeigorin Shusei, Second Edition (1872). The Japanese title reads “Waei (Japanese-English),” but the original title is A Japanese-English and English-Japanese dictionary. Figure 1 shows the entry of TARE ‘coll. Dare.’ Its explanation indicates that tare was standard pronunciation while the word was pronounced dare in colloquial usage. Images and electronic texts of the entire dictionary are made public on the Internet and available for anyone.
There are also old Japanese-language dictionaries and Chinese-Japanese character dictionaries. They are very important materials for linguistic study of Japanese language and its history. The oldest Japanese-language dictionary in Japan is Irohajiruisho compiled in the twelfth century. As indicated in the title, its entry is made in the order of iroha (traditional ordering of Japanese syllabaries.) The oldest Chinese-Japanese character dictionary in Japan is the Shinsenjikyo compiled in the tenth century. Over 20,000 Chinese characters are covered based on grouping by radical. It was possible to computerize the Japanese-English and English-Japanese dictionary mentioned first because it contains very few difficult characters. On the other hand, full texts of old Japanese-language dictionaries and Chinese-Japanese character dictionaries are not yet computerized because they contain many difficult Chinese characters and those in cursive style. Figure 2 is an example of an entry from the Shinsenjikyo. You can see a big “tare” character in the center line. Small letters are for the pronunciation and explanation of the meaning of the character “tare,” which are difficult to read.


What kind of research are you doing?

Hand-lettered books are called manuscripts and those written in the sixteenth century and before are called old manuscripts. The Shinsenjikyo mentioned above and other dictionaries have been handed down as old manuscripts up to present. The shapes of characters used in them are very different from those used today. Furthermore, because books which are several hundred years old are precious cultural assets, it is not easy to view the originals. This is why we use photographed editions true to the originals to decipher and study the originals. It is very important as basic study to decipher the originals and make them available for today’s researchers.

My laboratory promotes a project to create electronic texts of deciphered full text of the Shinsenjikyo and other dictionaries compiled in the Heian Period and build their database, which is named the “Integrated Database of Hanzi Dictionaries in Early Japan.” The abbreviation of the project is HDIC, with “H” taken from “Hanzi” (Chinese pronunciation for Chinese characters), “Heian Period,” and “Hokudai” (Hokkaido University).


What is your next goal?

Ability to decipher books written in letters different from those used today is essential for historical studies. Developing a method to learn this efficiently is a major challenge. Our task is to build a system that enables anyone having an interest in learning how to lead old manuscripts by comparing photos of illegible old manuscripts with their deciphered text.

Figure 3 Entrance of the Oriental Manuscript Headquarters of the BnF (2006)

Most old dictionaries in Japan compiled pronunciation, meaning and usage of Chinese characters. Because Chinese characters were brought to Japan from China, it is necessary to study old dictionaries compiled in China. A variety of dictionaries were compiled in China. Those dating back over 1,000 years still remain not only in Mainland China but also in Japan. Western libraries also have many such dictionaries. The British Library in London and Bibliothèque nationale de France in Paris, in particular, house a large amount of manuscripts including dictionaries written over 1,000 years ago and discovered in Dunhuang, China in the early twentieth century. Comparing these dictionaries with dictionaries compiled in Japan and casting light on how Japanese at the time learned Chinese characters are another task for us.