Japanese first-line spelling issues

Thu, 11/22/2012 - 12:54 — Haruo

I've started entering data from the current Japanese Protestant hymnal 賛美歌 21 [Sanbika 21]‎ and have run up against the ubiquitous issue of the Japanese writing system. This is most seriously an issue in the entering of first lines (and refrain first lines).

Japanese uses three distinct character sets, called "hiragana", "katakana", and "kanji". (In addition, there are competing systems for the romanization of Japanese words, e.g. the "official" system spells the mountain we call Mount Fuji as "Huzisan" rather than "Fujisan". I have used a more English-like romanization, called "Hepburn", in creating file addresses here.) Also, small (superscript-like) hiragana (or, more rarely, katakana), called "furigana", are generally written above any kanji that may be ambiguous, to direct the reader how to pronounce the word.

賛美歌 21 generally gives the first line (and Refrain First Line, if any) of the first stanza of each text twice, once within the musical staff and again beneath the music. But typically the appearance of the first line is different in each of these iterations. Taking SBK21 #3 as an example, the First Line is written

とびらをひらきて between the lines of musical notation

but

扉を開きて when it appears set as verse after the music, but with little furigana above the two complicated characters, reading "とびら" and "ひら" respectively. (When the same line is used as the Title, it is 扉を開きて without any accompanying furigana.)

All these variants would be romanized "Tobira (w)o hirakite", though there is no romanization in the instance.

It is common when typing Japanese to put the furigana in parentheses after the kanji they annotate. But this leads in some cases to absurdly over-parenthesized first lines, and of course the parenthesized furigana ruin the usability of the line for search-engine purposes... At the moment I have entered the first lines in this format, so #3 looks like this

扉（とびら）を開（ひら）きて

but I am wondering if it would be better to create a convention of giving Japanese first lines where this issue arises thus:

扉を開きて / とびらをひらきて or even 扉を開きて (とびらをひらきて)...

Haruo

Comments

Mon, 11/26/2012 - 09:50 — dianneshapiro

Japanese spelling issues

I don't know the answer to these questions. I would try posting to the Hymnary editor listserv (hymnary-editor@lists.calvin.edu). I think someone from the editorial board should respond.

Wed, 11/28/2012 - 16:00 — Haruo

Reply from Editorial Board via listserv; LOC romanization rule/s

I did, and Eric Stedfeld got back to me with this, which I am reposting here in case anyone here is following this topic but is not on the listserv:

I'm not sure I understand all the questions and issues, but I forwarded this inquiry to our Japanese subject specialist and cataloger here at New York University Libraries where I work, asking her for her input. Here is her response:
"Most of these questions depend on who the users are and what kind of searches they will do.

"I think it would be fine to list the title as it appears when it is listed as the title, e.g., 扉を開きて, without any furigana. In the context of a hymnal, I would imagine that the syllables are spelled out with the music to indicate what notes to sing when. In addition, it may be used to make it readable by everyone, including children. But I don't think those measures are necessary for this purpose. If hiragana is entered, a search engine will still find the kanji for tobira, as it will if the kanji is entered.

"Regarding romanization, I would urge the Hepburn system. That is what Library of Congress and many publishers use. The LC guidelines for Japanese romanization, http://www.loc.gov/catdir/cpso/romanization/japanese.pdf will probably be a real help to your colleagues. I am happy to answer additional questions, too, of course."

I hope this is helpful, or if there are other questions I can ask her please let me know.

and my response was as follows:

Thanks, Eric, and please pass my thanks to the NYU Libraries' Japanese specialist. Her reply is quite helpful, as is the LC romanization guidelines PDF she sent along. My active use of Japanese ended long before there were search engines, so I will take her word (which I see no reason to doubt) on their ability to find the right kanji based on kana. So my general standard will be to input the longest (most kana-laden) form; if a kanji is given with furigana I will simply enter the furigana as kana and assume the ability of searchers to be machine-led to the right kanji. Of course, if and when I manage to get page scans in there, much of this will be both clearer and nearly moot. And if and when full texts arrive they will probably be kanji unless I think there's reason to use hiragana instead (or, occasionally katakana, as in the kanji for "deep river" which are glossed in katakana-furigana as "diipu ribaa" in the Spiritual. ;-)