Electronic Databases at the National Institute of Japanese Literature (NIJL)

by Satoko Naito

December 2003




I here focus on the various databases available through the Kokubungaku kenkyû shiryôkan (National Institute of Japanese Literature) website.  Though several NIJL databases, including its currently most major project involving the Nihon koten bungaku taikei (NKBT) series, are still indicated as being in the development stages, they nonetheless provide reliable information important to scholars of Japan.  The databases are currently of most use to those in the field of pre-modern literature, but the NIJL has increased its efforts in developing archival databases useful to those in other fields including modern history.  Operated by the NIJL, these databases have the reliability and permanency lacking in some of the databases run by smaller institutions, and their contents will only expand in the future.  Yet, as these are all on-line sources subject to constant and numerous updates, this project should be used only as a guide and with the understanding that much of the information is inevitably subject to change. 


It should be noted that even considering the significant developments of the NKBT database, University of Virginia's Japanese Text Initiative (http://etext.Virginia.EDU/japanese/) is still considered a highly valuable source providing searchable full text reproductions of pre-modern literary works and should not be overlooked.  The site also houses an increasing number of modern texts.  Because it was developed primarily for use by English-speaking scholars, there are rarely any problems in retrieving the information on non-Japanese operating systems.  The database has recently greatly expanded and is very useful for its easy and accurate character and word search functions.  As noted by Satoko Shimazaki, however, the searches do need to be conducted in the form (kanji, kana, or rômaji) that it appears in the actual text, in contrast to the site's claim that searches can be performed in any form.  JTI includes, as of December 2003, 67 works including the Kokin wakashû and the Man'yôshû, and is constantly updated with new works.   

Kokubungaku kenkyû shiryôkan

National Institute of Japanese Literature

The NIJL was established in May of 1972 with the primary aim to preserve manuscripts and books relating to the study of Japanese literature.  Original texts and microfilms of these originals are stored in the NIJL, and an increasing number of texts are currently being digitally photographed.  The NIJL provides several useful tools for the scholar of Japanese literature, including a compilation of research published yearly in Kokubungaku nenkan.  It also has expended much effort to developing on-line tools in the form of the various electronic databases. 


Satoko Shimazaki provides information on the earlier stages of NIJL database projects in her guide to Full-Text Electronic Databases of Pre-modern Japanese Literature (http://www.columbia.edu/~hds2/BIB95/00e-texts_shimazaki.htm ).  As noted, some databases are still in the development stages, and some others are temporarily unavailable.  As of December 2003, there are seventeen databases in operation, and two which are temporarily discontinued.  This project provides an introduction to the accesible databases.  Though most databases are catalogues providing archival information of material housed by the NIJL, other databases deal with material located in other libraries and institutions, and there are some full-text and image databases as well.  Though some are similar in structure, the information retrieval process varies with each system.  For each database reviewed, included is background and general information of the project and introduction of the features of the database, including its various search options.  There is also information on obtaining membership for sites with limited access, and the future prospects of the databases in the early stages of their development.



1. Nara ehon database (image)

2. Nihon koten bungaku taikei database (full-text)

3. Kokusho kihon database (catalogue)

4. Bunken chôsa kâdo database (catalogue)

5. Kindai bunken databases

       a. Kindai bunken gazô database (image)

       b. Kindai shoshi database (catalogue)

       c. Meijiki shuppan kôkoku database (full-text)

6. Kokusho zasshisho mokuroku OPAC database (catalogue)

7. Maikuro shiryô wakosho mokuroku database (catalogue)

8. Oushû shozai nihon kosho sôgô mokuroku Union Catalogue of Early Japanese Books in Europe (catalogue)

9. Nijyûichi daishû database (full-text)

10. Kokubungaku ronbun mokuroku database (catalogue)

11. Shiryôkan shûzô shiryô database

12. Shiryôsho zaijyôhô kensaku sisutemu database 

13. Shiryô jyôhô kyôyûka database

14. Renga database (unavailable)

15. En'nô database (unavailable)


Nara ehon database


Image database

Open to the public


Seven ehon from the Nara period can be viewed in full color.  The available ehon are Karaito Zoshi, Sasayaki dake, Sizuka, Sumiyoshi Monogatari, Hioke no Soshi, Bunsho Zoshi,  and Homyodoji.  All volumes of each ehon are included.  There is a brief synopsis (in both Japanese and English) and product information for each work.  The database is very straight-forward; just click on the ehon desired, then click on any page to view in detail.   Every page, as well as both front and back covers, can be viewed in full.  The image quality is very good, and each two-page layout can be enlarged to a full-screen size. 


There is no information available regarding future prospects for this database.

Nihon koten bungaku taikei database



Limited Usage


Though the search-functions have been developed, the contents of this database have not changed since it was launched in April 1999.  As noted by Satoko Shimazaki's project (http://www.columbia.edu/~hds2/BIB95/00e-texts_shimazaki.htm#kokubunken), this database contains the 560 texts of pre-modern Japanese literature in the Nihon koten bungaku taikei published by Iwanami Shoten.  It provides the full-text of the works, along with character/word/phrase search functions and frequency information.  The speed of both database information retrieval and the search process, problematic before the database's most recent upgrade, is no longer an issue.  There also seem to be no problems with the display of Japanese characters on non-Japanese operating systems.  The restriction on the times that a user is able to access the database has also been eliminated, though the NIJL reserves the right to temporarily limit usage when necessary.  Despite the upgrades, the system is still in its 'experimental' stage: currently the NIJL estimates that the project will graduate from this stage in April of 2004, though it is unclear what aspects of the database will be changed after that time.  The various search functions seem currently to be fully functional.


As noted, use is limited to the research purposes of scholars studying Japanese literature.  Membership and password access is granted by email in as few as three days after request via an on-line application, and currently is still free of charge.  Incidentally, though the membership application asks for a 'user ID,' the NIJL issues a new, unrelated ID. 


The current database allows for several options in searching for a text.  Searches can be limited by period, genre, author, and title.  The periods are jôdai, chûko, chûsei, and kinsei.  There are twelve 'genre' that one can choose from, such as history, nikki and monogatari (which includes historical and military monogatari).  There is no limit to the number of periods or genre one can specify in the search; conversely, no specifications need be made, in which case the entire database will be searched.  The desired author must be entered in kanji, though the title can be entered in either kanji or kana.  In both cases, entry of multiple names/titles separated by spaces produces an 'or' search.  These four categories can be specified and used in any combination, so that one can search for setsuwa/shôsetsu in chûsei and kinsei, or search the entire database for works produced by Murasaki Shikibu.  The search function is in this way very user-friendly and allows for multiple approaches to finding a text, whether or not one has a specific work in mind.  The limitation of the author search entry in kanji, of course, is problematic for those who know only the reading of the author's name.  This information can be obtained in a number of ways (such as a simple Google.jp search, a search of Webcat, or of another database provided by NIJL, such as the Kokusho kihon database).   


After the search is conducted, the user is informed of the number of data conforming to the criteria entered.  Ichiran hyôji produces a new screen, which provides the text name, identification number, and number of characters in the text of each work that the search produced.  There are various links provided for each work, including one which leads to the text itself.  There are three HTML viewing options for the text; one can choose to view the text with or without line numberings, and with or wihtout kana readings.  For a small selection of works, there are also options to view the texts in XML, DTD, and SGML data formats.  The NIJL warns that the XML format may be very slow to appear in some browsers. 


Ichiran hyôji also provides links to three non-interactive, 'informative' screens.  These are 'detailed archival information,' 'information on original text,' and 'information on secondary text.'  The user will find most use, however, with the two remaining interactive links.  The Character Frequency Analysis page (Moji hindo bunseki hyôji page) allows one to view a list of characters in order of frequency of occurrence.  The characters can be limited to all characters (including symbols), kanji and kana, kanji, or kana.  One can also enter a certain character to find the number of times it appears in the text.  This function enables the user to determine whether a certain character occurs more frequently than another without having to scroll through the frequency lists. 


The last interactive link leads one to the Character Search page (Moji rei kensaku page), and is likely the most useful function of this database.  A specific character or series of characters can be searched within the text by Kensaku moji retsu, resulting in the page and line numbers locating the searched character(s), displayed alongside up to 100 characters (in total) that appear before and after it.  Kyoushutsu kensaki mori rei allows one to search for passages with two characters (or words/phrases) that appear 'together,' separated by up to 100 characters.  This allows for character search within a certain context.


As noted, the speed problems with the earlier versions of the database are no longer a factor, as the retrieval and downloading of information are extremely fast. 


Kokusho kihon database

Basic NIJL catalogue

Limited Usage


This database contains information on works in NIJL's Catalogue of Premodern Texts (Kotenseki sôgô mokuroku) but omits information regarding the location of the texts.  It is the compilation of the same information from Iwanami Shoten's 1989-1991 publications of Kokusho sômokuroku and the Kotenseki sôgô mokuroku (edited by the NIJL, published by Iwanami Shoten in 1990).   There are some minor differences in content between the mokurokus and the database, such as the addition of title and author kana readings unavailable in the printed editions of the Kokusho sômokuroku.   


The entries are useful for finding the various alternate names of texts and authors, and their kana readings, as well as for an authoritative list of works produced by specific authors. 


Similar to the NKBT database, the KK database is intended for use by scholars of Japanese literature, and thus membership is required for use.  As with the NKBT, applications for membership are available on-line and free of charge, though it is yet undetermined whether in its non-experimental stage it will continue to be so.  User identification and passwords are granted usually in less than a week via email after submission of application.  One point to note is that the user ID includes the word 'user' (followed by (currently) a four digit number): the ID in its entirety must be entered, not just the four-digit number.


Searches must be specified to either the 'works' or 'authors' databases.  The search can be conducted by work title, author, genre / era, or a keyword search can be conducted simultaneously in all fields.  In all fields, two phrases can be entered to perform 'and,' 'or,' or 'contains neither' searches.  The database can be searched for results to 'include,' 'match exactly,' 'begin with,' or 'end with' the entered phrases.  The 'ends with' search is achieved by leaving a hankaku space after the phrase.  In order to conduct a name search, there must be a backslash (/) separating the surname from the given name, and can be entered in either kanji or hiragana.  The year (seiritsunen) search should be entered as nengo and year (in kanji numbers), as this search function does not support Gregorian dates. 


The 'works' database includes the title and its kana reading, number of volumes, alternate names, genre, author name, publication date, while some entries also include annotations.  The 'author' database results include the name and kana reading of the author, dates of birth and death, vocation, alternate names and their kana readings, and works produced by the author.  Links of author names and titles allows for very easy movement from one database to the other.

Bunken chôsa kâdo database

Literature inquisition cards (card catalogue) database

Image database of card catalogues

Open to the Public


This project compiles information from cataloguing conducted by the division of document resources of the NIJL.  The information cards were all digitally photographed, and 'essential data' were incorporated into the database.  The catalogue includes material from various municipal and university libraries from around the country. 


There are four categories in which searches can be conducted, all of which can  'exactly match' 'begin with' or 'include' the entered phrase.  One can search by title, author/editor name, and storage location.  Either 'and' or 'or' searches can be conducted.  The search results can be displayed in sets of twenty or fifty per page.  The information includes title, library/institution or other storage location, and detailed information, including a scanned image of the actual hand-filled card catalogue used in cataloguing the books.  These display more specific information about such details as the size, binding, and color the text.    


Kindai bunken databases

Modern Documents Databases


Under Construction


There are three databases under this link dealing with Meiji and post-Meiji materials.  All databases are under construction in the early stages of their development, and much data is yet to be added.  Each database contains various information on modern (literary) documents. 


Kindai bunken gazô database

Modern Literature Image Database


Open to the public


This database contains scanned images of Meiji and post Meiji-era literature.  Though currently only one library has its Meiji era books on the database in full, the plan is to expand the database to include texts from various other libraries and institutions.  The Meiji and post Meiji literature housed at the NIJL is currently being integrated into the database.  It will not include the entirety of the NIJL Meiji collection, however, for copyright reasons pertaining to a selection of the works.  Covers and each page of the texts are digitally photographed.  Though there are several reliable image databases of pre-modern artwork and text available on-line, this database is currently one of few which deals exclusively with Meiji literature.  


As of December 2003 there is one participating library (Kôchishimin toshokan) though there are four more libraries (including, as noted, the NIJL) in the process of loading their images onto the database.  This library, founded in 1974 by a private donation, currently houses about 13,000 texts.  The bulk of the collection is post Meiji material.  The library began to collaborate on this database project with the NIJL in 1999.


The database is reached by clicking on the Kôchishimin toshyokan link, and appears as a three-way split screen.  The first screen on the left is an index, the middle a list of works in a particular 'genre,' and the right screen the table of contents of a specific selected work.  The index contains larger categories (literature, sôki, philosophy, history, etc) which are divided further into smaller genres (literary philosophy, poetry, nikki, etc).  The categories and genres shown in blue (with active links) have corresponding information in the database.  Currently, only Literature and Sôki have available links, and of these, only a selection of the 'genre' has data in the system.  Clicking on one of the links provides a list of works of that genre on the middle screen.  Again, of these works, only a selection is in blue (with active links).  By clicking on one of these works in the middle screen, one can activate the screen on the right to show the list of contents of the work.


From this screen, one can select a specific chapter of the work to jump to.  Images of every page are available, including both the front and back covers. The quality of the digital image is excellent, and the zoom-function is user friendly and of very high quality.  One can zoom with any focal point, though the images cannot be dragged as is possible with some other image databases.  These images cannot be downloaded or saved.   


As with any database in the earliest phases of construction, some of the links to the images do not work, and a Universal Viewing Error occurs.  Once the data is fully loaded, and once other participating libraries include their data, this will be an image database valuable to scholars of the Meiji and post-Meiji Japan who desire to see the original version of a printed text.   


Kindai shoshi database

Catalogue & Image

Open to the public


This database contains cataloguing information compiled by the NIJL with the assistance of individuals and institutions from around the country.  The catalogued texts are Meiji and post Meiji texts.


One can conduct a general keyword search, or specify the search in the categories of title or author which can be 'exact match' 'including' 'starts with' searches.  Though the title name can be entered either as kanji or with the kana reading, the other categories must be entered as kanji.  One can also search by publication date, either by nengo + year, or the Western (Gregorian) year.  A storage location (library or institution) search is also an option.  All of the above can be combined for 'and' or 'or' searches.


The results may be viewed by sets of 20 or 50 per page, or displayed in full on one page.  Clicking on a single entry provides a separate screen, with information regarding the text.  There is basic information (title and its reading, storage location), followed by more detailed information regarding the text itself (shoshi data), the structure of the book (shokei data) including size and binding information, and miscellaneous information (sono ta) including, for example, the color of the cover.  Under shoshi data, there is a link to kôsei, which includes information on all advertisements included in the original printed text.  Sono ta includes links to images of the text, though currently some of the links do not work.  There is also a link at the top of the page (along with the basic information) to the Kindai gazô database, but again, currently some links are not active.


Although the entries are not very easy to follow, there are numerous 'help' links that describe in detail what each category of information entails.      


Meijiki shuppan kôkoku database

Full-text database

Open to the public


The purpose of the Meiji shuupan kôkoku database project is to study the state of printing and publishing in modern Japan.  The database includes advertisements printed in serialized publications from the early Meiji period.  The database contains roughly 7,000 advertisements published mainly in Tokyo nichinichi shinbun from February of Meiji 5 to the end of Meiji 12. 


There are two search screens to choose from.  The basic search screen (Kensaku gamen) allows for specification by title, author, publication date (Meiji or Gregorian year and month), all as 'and' or 'or' searches.  One can also limit the search for advertisements published in journals or newspapers, or search the entire database.  The individual search screen (Kobetsu kôkoku kensaku gamen) allows one to search for a specific advertisement.  One can search by the publication the advertisement appeared in, or by dates (Meiji or Gregorian year, month, date).  One can also search the actual text of the advertisement, by a keyword 'includes' 'exactly matches' 'starts with' search.  All of the above can be combined for 'and' or 'or' searches.  Results can be displayed in sets of 20 or 50 per page, or all on one screen. 


The results display the basic information of the advertisements that fit the criteria.  Clicking on the ID link (Kobetsu kôkoku ID) from the basic search, or the Details link (shosai) from the Kobetsu search, opens a new screen with detailed information, including the full text of the actual advertisement.  There is no image of the advertisement available, nor does there seem to be any indication as to whether an advertisement had an actual image along with the text.


Kokusho zasshishozô mokuroku

Kokubungaku kenkyû shiryôkan kanranshitsu (OPAC) catalogue


Open to the public


This is a catalogue of NIJL books related to Japanese literature studies (Kokubungaku), including research materials, printed books, and Meiji era books.  The site also catalogues periodicals such as academic society journals, institutional/ academic bulletins, and literary coterie magazines.


Searches can be specified to include books, journals, or both, and keyword searches can be conducted by title, author, editor, publisher, or publication year.   They can be conducted in kanji, hiragana, katakana, or rômaji.  Search results include the above categories, and specify the storage location.  The registration number provided in each entry allows one to find the book.


The site also provides information on the hours of the OPAC Reading room.  Though it is to be used for academic research purposes, it does not specify whether the Identification required for entry needs to show affiliation to any academic institution.  For a fee, one can request texts to be photocopied.   There are also options available for those unable to visit the NIJL.  Scholars affiliated with universities or other academic institutions may request photocopies of the texts (for a fee) or, for a selection of texts, request an Inter-Library Loan.  The loan period, including the time for shipping to and from the NIJL, is 31 days.  Up to 15 books can be borrowed at a time.  Those individuals unaffiliated with academic institutions can also mail or fax a request (an application is available online) for photocopies of texts.  Though the photocopying service is expensive, and though it is unclear which books are available for ILL, these are services useful for scholars unable to visit the NIJL in Tokyo.  The ILL service, however, is unable to ship overseas; NIJL will in certain cases send books to Japanese family members or associates of the overseas patron, though the loan period is still only 31 days.  For inquiries regarding special cases, it is recommended that one email NIJL directly.


Maikuro shiryô wakosho mokuroku

Microfilm and Japanese-bound Books Database


Open to the public


This database contains archival information of 175,000 microfilm and 8,500 old Japanese-bound books of the NIJL collection. 


Searches can be conducted within the microfilms, the manuscripts, or both.  Both 'and' and 'or' keyword  searches can be performed by title, author, print date, or in all three simultaneously.  The search supports both kanji and hiragana, but the date should be entered in nengo, as it does not recognize Gregorian calendar years.  There are options to further limit searches by service kubetsu (specifying scanning/ photographing method) and storage location.  These can only be done alongside the main three category searches.

Results are listed in short format, and one can then specify specific entries to view its details, including title, author, type of book, and its storage location.  The cross-referencing of this index is very good; for instance, it is easy to jump from a work by a specific author to other works by same author.


Oushû shozai nihon kosho sôgô mokuroku

Union Catalogue of Early Japanese Books in Europe


Open to the public


This project began in 1988 under the direction of Peter Kornicki and Hayashi Nozomu following a conference held in Durham of the European Association of Japanese Studies.  Much of the cataloguing seems to have been the work of these two, though other individuals and librarians took part in the research as well.  As such, the number and scope of this project is currently somewhat limited.  The aim of this database, however, is an immense one, and is to 'include all books bound in Japanese style held in university and local libraries, museums and other institutions of all European countries.'

The countries appearing under 'List of Libraries and Other Institutions Covered by the Union Catalogue' are twenty.  However, of these Austria, the Czech Republic, Greece, and Portugal seem currently to be missing from the search function.   The 'List of Libraries' provides links to some of the museums and libraries from which the database acquired archival information. 


At current count, the countries that can be specified for search within the database number sixteen.  One can specify the country in which to conduct a search, or choose to search the entire database.  One can also display the entirety of books available in a specific country without entering a keyword. 


Though the site claims that word searches can be conducted in rômaji, kanji, or hiragana, searches using these various forms currently produce varying results.  A phrase entered in hiragana, in particular, does not produce a full-word search; the database considers the characters to be separate entities. The recommendation, then, would be to conduct separate searches using all three forms of writing.


Though there is some information on who researched which specific libraries / countries, there is no indication as to the future prospects for this database. 


Nijyûichi daishû database



Open to public


This was the first database to be launched as part of the full-text database project by the NIJL.  As indicated by Satoko Shimazaki, It contains the full text of 21 waka collections of the Shôho version Nijûichi dai shû. 


Search categories include poem title, author name, and annotations .  Most importantly, the full text can also be searched, in kanji or kana reading.  'And' and 'or' searches can be conducted between categories, or multiple phrases can be entered into one category to conduct 'and'/'or' searches.  Poems can be searched in a particular collection or in the entire database.  The Shin kokka taikan numbers have been added, as have minor annotations and information on kôi.   


Kokubungaku ronbun mokuroku

Japanese Literature Dissertation Catalogue


Open to the public


This database is a collection of dissertations regarding Japanese literature presented within Japan, and includes dissertations published in the Kokubungaku nenkan.  It currently includes data from Showa 16 to Heisei 13, and a portion of data from Taisho 1 to Showa 15.   Dissertations from Heisei 14 is currently in the process of being integrated into the database. 


One can conduct a search by title, author, or keyword in the 'Regular' search.  In a 'Detailed' (category) search, can search a dissertation by various methods, including by dissertation or publication title, author, reprint location, dissertation topic era, field, year of presentation, or a keyword search.  The keyword search cannot be conducted on the full-text of the dissertation.


Results include the Kokubungaku nenkan volume that it was published in, and the number of total pages.  The author's name is given in both kanji and hiragana. 


The following databases all deal with official documents that pertain to local history.  Documents deal mostly with the various histories of land-owning families, and exist as part of local governmental property. All databases are in the development stages and will likely undergo significant changes.  Following is a brief summary of what each site contains.


Shiryôkan shûzô shiryô


Open to public


When this project is finished, the user should be able to systematically navigate among several databases.  They will largely deal with official documents in the ownership of town halls, which can be searched on its own or can be found by following the index of the prefectures.  Unlike the other two databases below that will be used primarily to locate similar documents, this database also provides significant content information.  One can then have immediate access to extensive summaries of these documents.


Shiryôsho zaijyôhô kensaku sisutemu


'Limited' Usage 


Though the site is meant for restricted usage, currently the 'visitor's page' seems to allow any user to peruse the material.  Documents can be searched by 'old place name,' 'era' of content, or source.  It contains only the 'title' of the document with no information regarding the contents of the document, but does indicate the source.


Shiryô jyôhô kyôyûka database


'Limited' Usage


To enter this database the user must enter his/her name and an email address, along with a brief description of purpose of use.  Currently, it does not require membership or a password.  


This database, launched with financial support from the Monbu kagaku shô, deals with official documents. Currently, six libraries participate in providing archival information for the database.  


Temporarily Unavailable Databases


Renga database

As indicated in Satoko Shimazaki's project, this database had a comprehensive word-search function with options to limit the year, key words, etc.  There is no indication as to when this database will be available. 


En'nô database

There is no indication as to when this database will be available.