Japanese-English Kindle dictionary

[karunaji]

I couldn't find any JP>EN dictionary for Kindle so I decided to make one myself.

You can download it from: http://www.latvianforyou.com/ebooks/...Dictionary.prc

It is converted from the free EDICT dictionary that is available from JMdict/EDICT project. More info at http://www.csse.monash.edu.au/~jwb/edict_doc.html

No kana/kanji input is yet possible in Kindle therefore all headwords have been provided in romaji to enable quick lookup.

UPDATE: A new version has been made and can be downloaded. The changes are as follows:
- chōonpu (ー) now converted to duplicated vowel.
- some sentence examples added from the subset of Tanaka Corpus

[Alissa]

Originally Posted by karunaji

No kana/kanji input is yet possible in Kindle therefore all headwords have been provided in romaji to enable quick lookup.

It sounds interesting. I will try it later.

A question is who use this dictionary for what? I assume the primary use of a Kindle dictionary is to read a Kindle ebook. Hence, Japanese-English dictionary should be for English speakers to read Japanese ebooks. I believe, however, that most (if not all) of Jqpanese ebooks are written in Japanese characters (i.e., kana/kanji mixture), so Japanese dictionaries with romaji (i.e., Latin transliteration) headwords seem less useful.

I'm not criticising; I'm just curious.

-- Alissa

[karunaji]

My idea was to use it mostly as an stand alone electronic dictionary in cases when you have heard the word but don't know what it means. I used to have one on a Palm device but with Kindle being the main device now I no longer want to carry 2 devices with me. Kindle is enough and even better than Palm.

It could also be used by Japanese Kindle users who want to speak English but have forgotten the English word.

It could be made work as a pop-up dictionary for kanji/kana words as well but as it is mentioned elsewhere the problem with Kindle lookup is that it requires words separated by spaces. Japanese sentences are written without spaces so it will not work anyway. We will have to wait for Amazon developers to find a solution for this.

[odiakkoh]

Is Japanese and/or English your native languages?

[karunaji]

Neither. I am bad with foreign languages. That's why I need dictionaries

[odiakkoh]

Then I would steer clear of this one. Japanese is my native tongue (learned English in 1st grade) and I felt like I was reading a foreign language.

[karunaji]

It has nothing to do with me. As a dictionary Edict is primitive but it is virtually the only one available electronically. Please, let me know if you can find something better that I can use on Kindle.

I do not expect this dictionary to be a big success but I hope that its appearance will push publishers to release good Japanese dictionaries for sale sooner similarly to Kindle dictionaries for other languages.

[odiakkoh]

I do like Berlitz' language dictionaries but of course they don't have Japanese available in ebook form.

[AkumaTakeshi]

Thanks so much, was trying to find a Romaji to English kind of dictionary

Now I will try to find Genji Monogatari in romaji to try it.

Of course a kana/kanji one would be useful too but I'm not that advanced yet, I can't even separate the words, it's hard to know where one word starts and where it ends in kana/kanji

BTW for those who don't know, IMO Denshi Jisho — Online Japanese dictionary is the best online Japanese to English (and English to Japanese) dictionary, it uses EDICT too but it's well formatted, it accepts romaji or kana/kanji.

[Alissa]

Originally Posted by karunaji

My idea was to use it mostly as an stand alone electronic dictionary

Understood.

I downloaded and tried it on my Kindle 3. It worked nice. Thank you for the great work and sharing it with us.

I have some comments on the Romanization system used in your dictionary. I believe Romanization (Latin transliteration) of the headwords are not included in the original Edict and are your work. Please note and forgive me that I'm a native Japanese speaker and my view may be biased by my Japanese speaking experiences.

You use uppercase letters for (parts of) words written with katakana in the standard orthography. I felt some strangeness to see them. I know it is uncommon at least. I'm not sure it is good to use such casing in headwords, especially given that your dictionary shows hiragana/katakana/kanji notation of the word and it is clear whether a word is usually written in katakana or not. Also, I see no explanation that uppercase letters are for katakana words. I believe it should be explained in introductory text (e.g., in a paragraph beginning with "The traditional Hepburn rōmaji is used..." around the location 10.

You write in your introductory text as "The traditional Hepburn rōmaji is used with diacritics converted as follows: ā - aa, ī - ii, ū - uu, ō - ou, ē - ei (ee)", and use aa/ii/uu/ou/ei(ee) notation as explained in the headwords, while using ā/ī/ū/ē/ō in the introductory text (e.g., you wrote Hepburn rōmaji and not roumaji nor ro-maji.) I think it is inconsistent. If you prefer avoiding macrons, your introductory text should follow that convention. If you believe use of macrons is better, why not in the headwords? (I personally prefer use of macrons, BTW.)

You use an en-dash for a prolong sign (chōonpu 長音符). For examle, in your dictionary, a word アバンゲール is Romanized as ABANGE-RU. I don't think it is popular nor useful. As far as I know, all Romanization system for Japanese use the long vowel notation for a letter before the prolong sign. So, アバンゲール should better be ABANGEERU (or ABANGĒRU or just abangēru.)

You use no wordbreaking spaces between romanized words. As far as I know, all Japanese Romanization systems use spaces to separate words (or sometimes bunsetsu.) I believe the headwords are better breaked into some consistent units (words or bunsetsu or even keitaiso (morphemes)), especially given that Edict contains some long idiomatic phrases as headwords. For example, your dictionary now shows a headword for a phrase 彼方立てれば此方が立たぬ as achirataterebakochiragatatanu with no spaces in it, but it should be something like achira tatere ba kochira ga tata nu or achira tatereba kochiraga tatanu.

[Alissa]

Originally Posted by AkumaTakeshi

Thanks so much, was trying to find a Romaji to English kind of dictionary

Now I will try to find Genji Monogatari in romaji to try it.

Probaly you already know, but Genji Monogatari is writen about 1,000 years ago, and the Japanese language those days was very different from today's. Edict is for modern Japanese language. I don't think karunaji's dictionary helps to read Genji Monogatari...

[karunaji]

Thank you Alissa, for you detailed suggestions.

My romanization is simply an automatic conversion of the first available kana spelling variant. Thus I can't add spaces as there is no additional human input.

Macrons in headwords would be an issue because there is no way to enter them for search on Kindle. I could however add them as inflections to those who would like to use popup dictionary with macronized rōmaji text. Although I suppose that such texts are rare and hardly available for Kindle.

I will add an explanation that uppercase is used for KATAKANA and lowercase for hiragana. In practice it makes no difference as Kindle search is case insensitive.

I was thinking about how to deal with chōnpu. It is straightforward when using macrons but without them ē is represented both by ei or ee and ō by oo or ou. Which one do you think would be the best choice? Maybe I should make it consistent everywhere despite kana usage?

I was experimenting with kanji headwords and theoretically there is a way to make it work. One can select a part of the text, then press a spacebar instead of marking the end of the selection. Then in the search window remove the extra characters and select "dictionary" from the options below (hidden as the last element on the right) and a dictionary entry appears.

However, it is too much of button pressing and thus not practical. Instead Amazon could implement a special dictionary search for Japanese language that try would look up a string starting at the cursor and spanning for several characters. If no dictionary match is found for the string it can decrease the length of the string by one character and so on until match is found. It could even display several matches after a user presses ENTER.

[AkumaTakeshi]

Originally Posted by Alissa

Probaly you already know, but Genji Monogatari is writen about 1,000 years ago, and the Japanese language those days was very different from today's. Edict is for modern Japanese language. I don't think karunaji's dictionary helps to read Genji Monogatari...

I know but I couldn't think of anything else at the moment, if so, I'll just read some light novels to practice and extend my vocabulary. For now almost all my vocabulary is from watching Anime, lot's of Anime

[Aryn]

Thank you for the Japanese dictionary. Needed one badly for whenever I encounter all the unknown words when I'm reading. I can't get to the my Kindle to connect to the gprs as often as I like.
And this is fast and instantaneous!

[osnova]

What about Eijiro? It's much better than EDict.