Writings > Lexical Leavings


A couple of months ago, I wrote the following article about the new edition of Kenkyusha's New Japanese-English Dictionary at the request of the Japan Association of Translators. They have now posted it on their Web site in slightly different form. Other remarks by me about this dictionary are linked from the Green Goddess entry in the list of topics.

A Translator Tackles the Goddess

Tom Gally

In April 2002, I received an e-mail from an editor in the Dictionary Department at Kenkyusha Ltd. He knew about me from my posts to the Honyaku mailing list, and he wanted to know if I would be interested in working on the fifth edition of 新和英大辞典 (Kenkyusha's New Japanese-English Dictionary, known among translators as the Green Goddess, or GG).

As people who had read my Honyaku posts or looked at my Web site knew, I was already interested in dictionaries and had worked on several, including Kodansha's Furigana English-Japanese Dictionary and the Favorite Japanese-English Dictionary published by Tokyo Shoseki. Ever since I had bought my copy of the fourth edition of GG in 1986, I had been waiting for a new edition. I had heard a rumor around 1998 that an updated edition would be published soon, but that turned out to refer only to a CD-ROM version, which did contain some additional vocabulary but was otherwise identical in content to the 1973 fourth edition. So I was very excited to learn that a full-scale revision was under way and that I was being asked to work on it.

A few days later, I went to Kenkyusha's office and met with several of the editors. We discussed the project, and they gave me a floppy disk containing a text file for a short segment of the dictionary. The file had already been worked on by several Japanese editors. While some of the entries were the same as in the fourth edition, most had been extensively revised, and the entries for common words had been completely rewritten with new sense divisions and examples. Many new entries had been added as well. My job, and that of the other native English editors, was to correct and improve the English definitions that had been held over from the fourth edition, provide definitions for new entries, and translate the new example phrases and sentences into English.

Once I got used to Kenkyusha's system, the work on the text files went ahead fairly smoothly, though it never became easy. Unlike nearly all other types of translation, dictionary entries have essentially no repetition: after you have struggled with a particular headword and come up with good translations for it in various contexts, you're likely never to see it again until your next dictionary project, which may be years later or never. While sometimes I was able to apply the knowledge that I had gained from a decade and a half as a translator, more often I had to search through references and the Web to find definitions and examples that told me what the Japanese sentences meant and how I might translate them.

For example, one word I worked on was つのる 【募る】. The entry had two senses and a dozen example phrases and sentences that needed to be translated, including:
I came up with translations for these—you can check them on page 1742 of the dictionary—and then moved on to the next entry, which was:
つば 【唾】
Like the entry for 募る, this entry had been largely rewritten by one of the Japanese editors, so little remained from the previous edition. Among the new examples that needed to be translated were:
あ, 唾が飛んじゃった.
For the last one, I came up with "That's like 「spitting [《卑》pissing] into the wind." But I wondered: Should "pissing" be labeled 《卑》 for 卑語 or merely 《口》 for 口語? I put a note into the file asking the next English editor for his or her opinion. (Whoever it was must have agreed with me, for the label remained 《卑》 in the final version. How do you think "pissing" should be labeled?)

After 唾 came つば【鍔】, and after that was つばい ("a whelk; Buccinum tsubai"), which was followed by ツパイ ("a common tree shrew; a tupaia; a tupaya"), and soon after that were つばいもも 【油桃】, つばき【椿】, and つばさ【翼】. The work continued in this way, entry after entry, file after file, no two words the same, each word presenting new problems. Adjacent entries would jump from semiarchaic literary expressions to plant and animal names to onomatopoeia to medical terminology.

While the work was essentially translation, I could use few of the usual tricks of the translator's trade. Keyboard shortcuts or voice input would not have increased my speed, as I spent much more time thinking than typing. Unlike when translating in familiar fields or for long-time clients, I was constantly confronted with my own ignorance—about Japanese, about English, about science and history and architecture and fashion and every other subject that is described in words. I had to use all the resources at my disposal: dozens of monolingual and bilingual dictionaries, paper and electronic encyclopedias in English and Japanese, and of course Google.

The dictionaries that I referred to most were five kokugo dictionaries that I had on my hard disk: 新明解国語辞典, 岩波国語辞典, 広辞苑, 日本語大辞典, and 大辞林. I also frequently referred to the paper 三省堂国語辞典 and the 14-volume 日本国語大辞典. I used more than a dozen English-English general and learners' dictionaries from Britain, the United States, and Australia. In the middle of the project I switched my Web browser to Mozilla and set up a group of tabs specifically for the project. Clicking on the icon would launch four sites at once: Google and the for-pay dictionary sites of Kenkyusha, Sanseido, and Merriam-Webster. I also often used the Japan Knowledge and Encyclopædia Britannica sites. To avoid plagiarizing, I rarely referred to other Japanese-English dictionaries except for those published by Kenkyusha, specifically New College and Luminous.

After ten months or so, I was shifted to answering questions from the Japanese editors. At first they sent them to me mostly by e-mail, but as the project neared the end I started going to Kenkyusha's office once or twice a week to deal with questions face-to-face.

There were two main types of questions: Those asked by other native English editors while working on the text files, and those asked by the Japanese editors. The British editors, for example, asked many questions--or just gave up altogether—about the appropriate English translations of Japanese baseball terms, and I and other Americans familiar with baseball were asked to answer them. The Japanese editors would catch errors or omissions in the drafts of definitions or translations and ask for corrections. Many new entries and examples were also inserted at this stage as the Japanese editors noticed words and collocations that should be added.

The entry for さんか 【参加】, for example, included this phrase example:
参加を呼びかける call for sb's participation
An editor noticed that the English referred to a specific somebody, while the Japanese could also refer to any number of unspecified people. The entry was revised to
参加を呼びかける (個人に) call for sb's participation; (不特定多数に) ○○.
and I was asked to provide a translation for the ○○. (I proposed "(put out a) call for participants.")

An editor found the new horseracing term さんれんふく 【三連複】 and wanted to add it to the dictionary. It was defined on the Web as "順不同で1着から3着までの馬番を当てる馬券 (2002年7月13日より発売された)." I had no idea about the appropriate English, but through some Web searching I found the term "trio" used in that meaning at tracks both inside and outside Japan. For the entry, I proposed both "trio" and an explanatory translation, and that's how it appears in the dictionary:
さんれんふく 【三連複】〔競馬〕 a trio; a bet on the top three finishers in a horse race in any order.
In the final months, thousands of such questions—none of them easy, each presenting unique problems—were fielded by over half a dozen native English editors both in Japan and overseas.

When I went to Kenkyusha's office to answer questions, I would also read galleys of the laid-out dictionary pages. I typically made several changes per page. Most of those last-minute corrections were minor—removing the symbol that means "British English" from an expression that is also used in American English, for example—though I did catch some misspellings and mistranslations, and I made other changes that I thought would improve the dictionary. Toward the end, many additional Japanese editors were brought in from other projects to help finish up GG; it was impressive to walk through the office past row after row of desks and see every person reading galleys of the same dictionary.

One Saturday when I spent the entire day at Kenkyusha, there happened not to be many questions, so I devoted almost all of my time to reading galleys. In about seven hours of work, I managed to read ten pages, meaning that it would take a year of full-time work for me to read the entire dictionary.

As the final deadline neared, the core editorial staff went to the printing plant in a suburb of Tokyo to finish up the proofreading, and I volunteered to go along on the last day. We spent most of the day reading galleys and confirming the very last corrections, but after lunch we were also given a tour of the printing plant. We got to see our handwritten changes being input into text files in a special markup language; the pages were then laid out on UNIX workstations, printed out on film, transferred to large aluminum offset sheets, and printed on huge presses. On the day we were there, the presses were turning out 32-page sheets from the already-completed と and な sections. We finished up the proofreading around five p.m. with a round of applause and a big sigh of relief that the editing was finally finished. I had been helping with the project part-time for fourteen months; the others had been working on it full-time for as long as eight years. The completion of the fifth edition followed that of the fourth by twenty-nine years.


My name appears in the finished dictionary as one of the 編集委員 and 執筆者 and I will receive royalties on its sale, so the following assessment should not be regarded as unbiased. But I will try to be reasonably objective.

Clearly the fifth edition of Kenkyusha's New Japanese-English Dictionary is the largest and best Japanese-English dictionary ever published. The only contemporary dictionaries that come close to it are Eijiro and Sanseido's Grand Concise Japanese-English Dictionary. While both can be useful, the Japanese-English side of Eijiro seems to have been largely backwards-compiled from its English-Japanese counterpart; it misses segments of Japanese vocabulary, and it often offers inappropriate translations for Japanese words. Grand Concise is padded with many unnecessary entries from its English-Japanese version, and it would have benefitted from more input by native speakers of English.

The strongest point of the new GG is its example phrases and sentences. While some examples have been carried over intact from the previous edition, most were newly written or revised for the new edition by native Japanese editors chosen for their knowledge of and feeling for Japanese. It is not easy to create short yet typical and illustrative examples of word usage, and many times I was struck by the naturalness of the Japanese examples. The examples were translated by native speakers of English, and our translations were checked for accuracy by Japanese editors. As one of those editors remarked to me, thanks to its wealth and variety of examples, the dictionary is valuable not only as a look-up tool but also as a book to be browsed and read at length.

All of us who worked on the dictionary know, however, that it can still be improved. One problem that nags at me, though few readers are likely to notice it, is inconsistent spelling. While we referred to a particular English-English dictionary when more than one spelling of a word was possible, hyphenated and compound words in particular—such as "year-end" and "yearend," or "bestseller," "best-seller," and "best seller"—sometimes slipped through the cracks and are spelled differently in different entries.

A more subtle issue is the variation in the register and style of the English in the example translations. The English editors were natives of several different countries and ranged in age from around thirty to around seventy. British and American variants are noted in many cases but not in all; some translations, especially of conversational examples, have a clearly British or American feel, for it is in the informal registers where English dialects vary the most. Some editors also seemed more conservative in their choices of expressions, while others were more likely to include recent colloquialisms and slang. My own translations, I think, tended toward the informal side, though I once changed the usage label for the intensifying adjective "fucking" from the previous editor's 《口》 to 《卑》.

While GG now contains some 480,000 entries and examples, it can of course be expanded even more. The new edition includes many words and expressions that are listed in no other Japanese dictionaries, but by attentively reading magazines and other contemporary texts one can easily find fairly common words and meanings that do not appear even in GG. I did a little word-hunting for GG myself, and a few words that I found were added to the fifth edition, including ITゼネコン, 串刺し計算, 毛玉っぽい, 注力, 電解水, 握りこむ, ヘアメーク, 保定, 誘眠剤, and レス. In the months and years ahead, I hope to help collect Japanese words and expressions that can be added to digital updates and an eventual sixth edition. I also hope that the sixth edition will appear sooner than twenty-nine years after the fifth.

(September 22, 2003)