I think in the end we’re simply talking about the initial learning of the word. So using some simple examples, 刀 has a pretty simple, straightforward meaning, it doesn’t take much to understand what that is. However a word like かける what you are learning is simply the initial step, whether you learn it in a list or in a sentence, you’re going to hear it in ways you don’t understand later one, no matter what. So repeated exposure is what’s most important.
Side note, Japanese definitions tend to be superior as they tend to be an explanation not a single word translation
I think you can learn words from lists for as long as you like. I still do from time to time and I can read a NHK easy article and understand it (definitely not bragging, just putting things into context).
The “reason” not to do it I guess is cause it’s just less efficient (it is said) since learning words from context allows for you brain to make stronger connections to the word. If I study a word from a JLPT list (which I do sometimes) I learn it as “just a word” and I don’t really have a strong real life connection to it. When I read Death Note and learn 心臓麻痺 = heart attack, it is way easier to remember cause it’s context is ingrained in my brain and I can pull information from scenarios and imagery.
text hooker (I think more people would be into learning Japanese if they knew about “text hookers”, lol)
VN
I think having a tool that could identify the most used words of a text could be useful. I wonder though, maybe it would also be useful to know which are words less commonly used in general before tackling a text.
It is a form of extracting the text from something like a visual novel. Most instances I’ve heard of involve attaching it to a visual novel to extract the text and paste it either on something the test hooker has or a blank page.
I’ve been trying to switch to a j-j dictionary lately and can recommend this. Even if you don’t understand everything in the explanation, I still think it’s better than an English translation. It also stays longer in my memory even without SRS.
That’s true, but there will be a time whenever you’ve likely learned next to all of the words in a given medium. Like, you could probably know next to all the words in shonen manga but not be able to read anything pre-WWII Japan for beans.
I think that reaching that point is a good goal for most people asking this type of question.
Wew, this is one of my passion subjects. I get pretty hard about efficient learning.
Is that CB’s analysis tool or whatever it was called?
Vanilla’s method is good. As long as you’re reading in a format that has copiable text you’re pretty much gucci. Otherwise if you run into a kanji you don’t know the reading for it turns into a major pain.
As for my personal methods: I never read manga because the vocabulary is too limited and the time it takes to find a new word is generally too high. I usually like to read children’s books (Aoi Tori Bunko, Tsubasa, etc.). The first time I read I made the mistake of trying to use sticky notes. It was absolutely hilarious because by the time I finished there were so many that the book wouldn’t even close.
After that I decided I’d just go through the whole thing on an e-reader and highlight/mark all the words I don’t know without looking them up. When I’m finished with the book I go through all the marked words quickly and add them to an SRS. It’s a lot quicker than trying to flip through a 200 page book.
Here’s what the e-reader method looks like in practice. You got a list of words and if you need the context you can just click and you’ll jump to the sentence it’s in.
Lately I’ve been using a website I’ve been programming. It’s kind of like the thing Vanilla mentioned but it shows the words in order and filters out words you already know on said site or on Wanikani. You can add words from books straight to an SRS so it’s basically the method I described above except you can find out the words you don’t know in advance (shameless plug).
I’m curious about that website .
I posted a while ago about some text analysis tools I found. I was planning to use them when I start with ebooks. But meanwhile I’ve been useing them to grade the shows I watch, using the subtitles files. The results have been… not so great.
Yeah. I use it because I can also automatically filter out the words I learned on WK. The “high frequency” words can be pretty eh since it includes a lot of particles and stuff I already know, but it has proved pretty useful for some vocab.
For instance, one of the chicks in this vn is into aquariums and marine life to put it simply. When I opted to go down her route, I just fed her route script though that program and learned all the marine biology terms that I knew would be coming up a bunch.
Text hooker is just something that extracts text from a program/window on your computer. By extracting text from a visual novel, I can feed it through a parser/dictionary so I can easily look up any word I come across. Me playing without a text hooker is pretty much synonymous with me playing without a dictionary.
VN is a visual novel. xyz pretty much summed that up
Yeah that’s definitely a weakness it has. I copied and pasted the first 1000 words of this (which includes particles) into the ‘ignore these words’ file so that there’d be less trash in the output. Also it looks super ugly which has always bothered me.
I’m still pretty new to learning japanese so my reading comprehension is pretty low. I’ve been using white rabbit graded readers Japanese Graded Readers Level 0 Vol. 1 – OMG Japan which have an audio portion and you can guess words based on context and picture.
I would also recommend the android app TangoRisto http://www.tangoristo.com/ . It provides a list of simple reading material like nhk easy ranked by vocab difficulty. You can tap words to show an english definition. It also lets you show or hide furigana based on jlpt level.
I’d say even simply Google Translating a dictionary entry can give you a better idea than something like EDICT. Of course, then you have the words with 50 definitions.
Even then, you just need to read a Shonen about a different topic and you’ll suddenly come across a lot of words you’ve never seen before. Since it’s on the top of mind right now, Hikaru no Go is filled with a ton of words you’ve probably never seen. Or words you have seen, used in a different way.
I started out using context sentences for vocab, and I think it was a really good method when I was still not at a high enough level to read anything but context sentences.
When I started reading for real I also started adding every word I look up (how much I actually look up varies depending on whether I feel like reading extensively or intensively) to anki, and then it just felt like such a pain to have to read a whole sentence for each flash card, so for the past year and a half it’s all single vocab. You can get through so many more words in a day like this, and I’ve found it to be a very efficient way of building vocab.
Like @LaughingLiving says, since you’re now doing actual reading, the nuances will come from seeing the word in context.
Recently I’ve also started looking up the words in a j-j dictionary as a first resort, so even if they don’t have sample sentences, my anki cards are now starting to have answers written in Japanese.
I agree with @sigolino and @Syphus that an explanation is better than a translation. And of course there’s also the obvious benefit of simply being exposed to more Japanese.
For grammar though, sentences seem to be the only thing that sticks for me, even if I think it makes me cheat by learning the whole sentence.
This whole time I thought visual novels were actual physical novels with an anime picture here and there to illustrate the important scenes, this is slightly disappointing…
Mostly they’re right now too time consuming. And if using a Mac, it requires to move back and forth a lot with a VM (for me at least).
The cb text analysis tool It’s great since it can analyse a folder full of subs files. It also provides some personalized reading difficulty report.I’m not shure how useful are the Readability Reports it makes but the Readability Report based on user’s known vocabulary sounds great.
The unnamed text analysis tool can only work one file at the time, and doesn’t give you the accumulated percentage of known vocab (I’ll be aiming at 80% vocab before watching a show). But this one works on Mac just fine.
Anyway I think they are great tools, but I’m just at the point that I have to create an accurate list of my vocab (specially one that includes particles) to get the most of them. And that part it’s proving to require quite some time.
Does your tool work in a similar way? Will it work for Mac??
Edit: I think I’ll make a post with examples once I get to make a good list so the accuracy problem it’s less of an issue.
.