Tips on Interactive Searching

(Note: Displaying the Japanese characters below requires software that can read and input Japanese characters.)



Enter Japanese characters or English (or Romaji) in the search form, or combinations of Japanese and English. Example: to search for 秋風 in Kokinshu, in the form entitled "[Further constrain by] ... title," enter "kokin wakashu" or "kokin". Then switch to Japanese character entry, and in the form entitled "Search for word or phrase," enter 秋風. Click on "Submit Query" to get a keyword list of the 18 occurrences of 秋風 in Kokinshu. Click on a keyword to see the full context -- for example, poem no. 234:

女郎花ふきすぎてくる秋風はめには見えねどかこそしるけれ
Some Japanese clients like Unionway/KanjiKit by default put a space after each kanji. Our search forms do not recognize kanji with spaces between them. If you enter 秋 風 [with a space], the search will tell you that 秋 風 does not appear in the Japanese texts. If you enter 秋風 [without a space], the search will find the correct instances of 秋風 in the texts. Most clients allow you to change the default so that spaces are not inserted between characters. As an alternative, you can manually delete spaces between characters.

If you click on "compound search," you can search for one character or character string or word or words in Japanese or English or Romaji, near another character or string. Example: Enter a search for 秋 within 40 characters of 風. Then switch to English input and constrain the search just to "kokin." You should get the answer that there are 24 instances of 秋 near 風 in Kokinshu (including the 18 秋風 above). An example of 秋 near 風 is poem no. 781:

吹きまよふ野風をさむみ秋はぎのうつりも行くか人の心の
Searching in Japanese characters requires some creative testing of different characters. For example, in addition to 18 occurrences of 秋風 in kanji in Kokinshu, in the JTI edition of Kokinshu there is one occurrence of あきかぜ in kana. "yama" can be either 山 or やま. It is safest to search for both kanji and kana equivalents of your search term(s).

In compound searching, you can set the range within which you want to search for x near y to 40, 80, or 120 characters. However, an appropriate SGML division of the work being searched determines what part of the work is displayed when you click on an item in a keyword list. For poems such as those in Kokinshu, the division is the complete poem with author, commentary, and the poem. For prose like Hojoki, the division is a paragraph. Thus, even if your compound search identifies 秋 in one Kokinshu poem near 風 in another poem, you can display only one of the poems at a time, or else display the whole section in which both poems occur.

Highlighted Search Terms: The characters or words that you search for are now highlighted in red in the search results. Highlighting works only if you have cookies enabled in your browser. To enable cookies in Internet Explorer 5.x, go to Tools/Internet Options/Security. Select Internet as the Web content zone. Click on Custom Level. In the resulting Settings list, scroll down to Cookies, and click on Enable. To enable cookies in Netscape 4.x go to Edit/Preferences/Advanced, and click on Accept all cookies.

You can arrange the display of search results to show each instance of a search term, or you can group the instances by works. By default, search results are displayed for individual instances. At the bottom of the search page in the pull-down menu for "Group results by [match | work]," click on "work" to display results grouped by works. For example, to see all the instances of 秋風, search for that term by "match." For the texts available in September 2001, you will find 240 instances of 秋風. If you select to group results by work, you will see that there are 18 instances in Kokinshu, 8 in Saigyo, etc.

Japanese Text Initiative
Electronic Text Center | University of Virginia Library
PO Box 400148 | Charlottesville VA 22904-4148
434.243.8800 | fax: 434.924.1431
Etext Home | Library Home | Search the Library Web
Maintained by:
Last Modified: Tuesday, August 31, 2004
© 2004 The Rector and Visitors of the University of Virginia