One of the problems of studying Japanese is the lack of vocabulary. You can study lots of grammar rules, or even kanji, and still be unable to read a single piece of news from a newspaper because you just don't know the words.

But, what words "should" you know? That's a difficult question... there are grequency lists, list of words according to each Nouken test level (JLPT), but its still hard to chose the words that will be the most useful for each one.

Well, I had an idea a few days ago, I haven't implemented it yet, but thought that it was interesting enough as to post it here, just in case that anyone would get benefited from it.

1.- Create a list of vocabulary using :

A: Perapera-kun
If you don't know Perapera-kun, it is a Firefox extension that uses EDICT. When you hover over a japanese word, it gives you the reading and meaning (among other options). There is an option that if you pres CTRL while a definition is being displayed, that definition will be stored into a text file. Be careful, you have to save the info in "tabs" and not "EDICT" format.

And/or
B: Wakan
This one is a tool for learners of Japanese that includes a dictionary, lookup tools for kanji, and the option of creating and exporting vocbulary.

2.- After some time has passed and you have enough words stored into your text file,
open it with a words processor. Chose the right format, so japanese characters don't turn into mojibake.

3.- Copy the text from the word processor and paste it into a spreadsheet.

4.- Copy the column with the words (not pronunciation nor definition, just the words) from the spreadsheet and paste them into a new text file.

5.- Use a program like TEXTStat.
It is a program that (among other options) can create a frequency list based on your own text files. So create a frequency list of all the words you have looked up with Perapera-kun.

6.- Feed your SRS with those words, you can use or give priority to the ones with more frequency if you want.

This way, you won't waste time studying words you don't use, and you would learn faster the words you need the most. So, it would help you to understand stuff you read frequently, like newspapers, blogs, fora or whatever you use to read, as each type of text has its own vocabulary.

You can use Perapera for websites (as it is faster than Wakan on that context) or Wakan if you want to extract vocabulary from other places. Wakan gives you the option of creating different vocabularies at the sime time, while with Perapera you can use only 1 text file for all your words.

Any comments?

hmmm... I have the opposite problem, I find vocab is easy to come by, but grammar and context for the vocabulary is more difficult.

Interesting. Finding useful vocab can sometimes be really tricky stuff. I've found frequency dictionaries of German and other languages, but I've never been able to find a frequency dictionary of Japanese. (I gather that doing the research to compile it is the tough part.)

I did find a file somewhere on Jim Breen's site, I think, that is a frequency list of Japanese words in newspapers over a four-year period. The problem is that it's only words used in newspapers, so it's not as useful as a frequency dictionary that would list the 3-5,000 most frequently used words across all media.

http://ftp.ftp.cc.monash.edu.au/pub/niho...freq_ck.gz

You can view the resulting file in Firefox once you gunzip it.

Personally, but then again I live in Japan, I keep a notepad/pen in my pocket and write shit down all day. Well, by all day I mean "once in a while" but whatever.
That or when I play games I write down words I don't know. Not everything, or I'd get pissed off. But when I really have NO IDEA what the sentence is about I write down a word.
Grab sentences out of OS X's dictionary at a later time.

Weighting one's own vocabulary is an interesting thought. I'm not familiar with TEXTStat, but it sounds interesting. I just wonder how one would decide on which words one reads all the time and which one's you don't. I don't write much in Japanese if at all so automatically generating a frequency list from my own files would be meaningless.

But if you loaded in a really large file from an electronic book and the TEXTStat software could scan that single file for frequency, that would be pretty cool. Then as you read the book you could be simultaneously studying the vocab with SRS. Might be worth looking into.

Cheers.

I agree that it's not the vocabulary, it's the context we're missing. Vocabulary you can get at most JLPT sites (even Anki has a flash card set), but you won't see the way to use them in sentences.

Luckily, Fabrice will be answering our problems soon enough. We'll have sentences and vocabulary coming out our ears. Cannot wait.

Manga is especially great for building vocab. Some comics have furigana over every kanji. In addition, the dialogue in manga is written in a conversation style that is exactly the way you'll hear many people speak in Japan. It's best to find a manga that most closely matches the style and vocabulary of speech that you're interested in.

My advice is to use a UV blacklight pen and highlight words or phrases you don't know. Since the ink is invisible, you won't see the highlighting on the page. I find that this gives the least distraction when reading. When you're done with reading, just flash a UV light and the everything highlighted will glow. Enter the vocab into an SRS and you're set.

For me, this works really well, since I can often recall scenes where certain vocab words were used. So when I'm talking to people, I have flashbacks to specific scenes in manga and I just act like the characters in those scenes (within reason :-)

Just working my way through a textbook plugging vocab into Anki is my preffered method. All the vocabulary comes in context and is always varied (it won't use 食べる for every food vocab you encounter). With abstract nouns and verbs, I can easily include example sentences to clarify to meaning.

I'm working through the 漢字学習ステップ books. The sentences aren't long, but holy hell are they many and varied. Common word context is simply getting pounded into my head. I highly reccommend it.

Pen and pencil, or the save feature on your dictionary during the day is great for people in Japan.

I happen to do quite well with vocab lists, but spend the time to discover the nuances of the usage before I start drilling them.