wrightak Wrote:The raw data is freely available on the google spreadsheet. You can watch me working through it. Just click on File->Save As to download. Then create your own model in Anki, manipulate the spreadsheet such that each field in the model represents each column, then highlight the data, copy, paste into notepad, save as UTF-8 txt and import to Anki.

Wrightak, I'm unclear of what to do with the spreadsheet, as it has quite a few more columns than your original Anki deck model you posted on your Google pages. Should I combine the Word columns together and the Hiragana columns together?

to_nihon Wrote:Wrightak, I'm unclear of what to do with the spreadsheet, as it has quite a few more columns than your original Anki deck model you posted on your Google pages. Should I combine the Word columns together and the Hiragana columns together?

The first thing you need to do is create a model that you like. Perhaps you could use the existing Heisig model as a base. Here's a really simple example.

Model name: JRTK
Field1: Frame No. (example entry: 16)
Field2: Kanji (example entry: 古)
Field3: Word (example entry: 古い)
Field4: Reading (example entry: ふるい)
Field5: Sentence/Phrase (example entry: ふるいお寺)

Card1:
Q: Field4
Field5
A: Field1
Field2
Field3

In order to input data into Anki, you need to make a text file where each line looks like this:

16　　　古　　　古い　　　ふるい　　　ふるい本

i.e. a text file where the each line contains values for the fields, separated by tabs. (I think commas and semicolons etc. are also supported now)

The easiest way to make a text file like this is to arrange the spreadsheet so that the first column is field1, the second column is field2, etc.

You then highlight the data, copy, open Notepad, and paste in there. Then save the file as UTF 8 txt and import into Anki.

That make sense? Do you know how to manipulate the spreadsheet to get what you want? Highlighting columns, cutting and pasting them, etc.? For cases where there are empty cells, you can use a formula like "=IF(C1="", D1, C1)" which will take the value in D1 if cell C1 is empty. That way you can take the value for Word2 if the Word1 column has an empty value.

The spreadsheet is for assembling the data but you don't want to put all of it into your deck. Just put what you want in by jigging columns around and highlighting the bit you you want.

That makes sense. I figured most of that out last night but didn't know that formula. Think I'll use that too.

Are you planning to go to 3007 or just stick with what you have? I imagine it would get tougher if, like it says, many of the 2043+ are used only for names.

I've been adding a few kanji from RTK 3 to my personal deck but none of them are in the spreadsheet. Unless they introduce some funky new primitives, I think that's the best way to do it - to add them as you encounter them.

Kanji for names are pretty annoying because you see them so infrequently and it's hard to get a good feel for their meaning.

cangy Wrote:I've just been playing around with the JLPT vocab lists and have built a list of all vocab indexed by RTK1 kanji that might be useful as Japanese prompts for an RTK1 Anki deck

here's a new version with furigana and including core2k: http://ichi2.net/anki/wiki/ContribFugoun...anji-vocab

For words with two (or more) kanji I would like to add furigana so I can tell what the reading of the kanji under review is. For example for question わがはい the answer is 吾/吾輩 but I can't tell if the reading for 吾 is わ, わが, or わがは

In Anki what is the easiest way to have the answer show the kanji word with furigana?

I've tried renaming the 'Kanji' field to 'Reading' and having it be 吾輩[わがはい] but Anki doesn't display furigana just 吾輩[わがはい]

Thanks!

For わがはい,
吾:わが
輩：はい
(You could find this out by investigating with Rikaichan and looking at the kanji info)

Do note that 我輩＝吾輩 are practically interchangeable. It may depend on the period of writing. I would have expected the first, in my opinion.

As for getting anki to display furigana, you need the "Japanese Support" plugin, I think. Either way, I'm pretty sure you should direct anki questions to the anki forums and not here.

dat5h, you didn't get my question. I'm not asking how to find out how to read a kanji compound. I'm asking how to use wrightak's Anki deck and have it display furigana for compound words.

This is definitely an Anki question but considering the deck was started on this thread and that most of the people on this thread are probably using it I thought it appropriate to ask here first. Someone on this thread might have already a solution tailored specifically for wrightak's work.

PS And yes, I already have "Japanese Support" installed

Not sure about furigana in Anki, I've avoided using it so far.

The new deck I will release will change the cards as follows:

Q: わがはい（わが輩）
A: 吾輩

So that will solve your problem, even if it doesn't use furigana.

I find using furigana tough in Anki and I even get the feeling that sometimes there is no way to use it exactly as one would like.

Youre new deck will definitely solve my problem. Looking forward to seeing it!

totsubo Wrote:In Anki what is the easiest way to have the answer show the kanji word with furigana?

I've tried renaming the 'Kanji' field to 'Reading' and having it be 吾輩[わがはい] but Anki doesn't display furigana just 吾輩[わがはい]

Thanks!

see the post directly above yours and also the anki faq

tl;dr: tag the model with "Japanese" and maybe put a space before the kanji in the reading field

Wrightak, thanks for all the great work!

I want to start RTK with Japanese keywords, so I was taking a look at the spreadsheet to construct some of the spreadsheets while the full deck is being completed. However I have a question on the process, specifically in which order are new cards being "updated" and how many have been updated in the master spreadsheet? Basically the question is "Which cards are 'safe' to use?".

I don't mean to be rude! Your work has at least one very anxious fan! =D

I don't know if this is a valid question here, however as I just finished going through the RTK 1 , I am starting to look for decks to start studying the way Wrightak has done.

In a couple posts above, Wrightak noted there will be a new deck released. Could I get my hands on that deck? I'd love to start as soon as possible. If it is incomplete, let me know, perhaps I can lend a hand.

Thanks either way!

Hi,
why are you not using for 自 the keyword 自ら?

ok, 自分 is wellknown, but with mizukara you learn a new word, too.

Teskal Wrote:Hi,
why are you not using for 自 the keyword 自ら?

ok, 自分 is wellknown, but with mizukara you learn a new word, too.

I think the idea is to use kanji for words that are well known, and/or match the original RtK keyword as best as possible. So as far as new words you want to learn, you could make your own cards/decks for that. Of course there's going to be a lot of new words to learn with the JRtK deck, but that's not the main objective... I think.

I haven't tried it yet, but I really want to. It seems wrightak is away though

rogerdodge Wrote:I don't know if this is a valid question here, however as I just finished going through the RTK 1 , I am starting to look for decks to start studying the way Wrightak has done.

In a couple posts above, Wrightak noted there will be a new deck released. Could I get my hands on that deck? I'd love to start as soon as possible. If it is incomplete, let me know, perhaps I can lend a hand.

Thanks either way!

I'm kinda holding out for this too before starting with Japanese Keywords... any news Wrightak?

You could always bug "wrightak" on Twitter. ;p

nest0r Wrote:You could always bug "wrightak" on Twitter. ;p

I proudly have never visited Twitter. Ever! However it's tempting now...

Twitter scared me... and I'm young.

Somebody else will have to bug Wrightak!

In my Anki deck, I've started adding Japanese keywords as a supplement to, rather than a replacement for, Heisig's English keywords. It helps straighten out confusingly similar English keywords (like town/village), but since I still have the English keywords on the cards, the Japanese keywords don't have to be unique (for example I'm using は for both "blade" and "tooth"). (Using sentences would also work, e.g. はをとぐ vs. はをみがく, but I'm trying to minimize the amount of reading and additional vocabulary required to answer the card.) Now I feel like I'm learning real Japanese words, whereas when I was using only the English keywords I felt like I was putting that off until after I'd learned all the kanji.

Following an exchange earlier in the thread: http://forum.koohii.com/showthread.php?p...7#pid66907 and while we're waiting for wrightak to stop having a life and come back here with updates...

Why not use the RTK Lite KO2001 template (http://forum.koohii.com/showthread.php?tid=993) and the smart.fm-in-KO2001-order list here (http://forum.koohii.com/showthread.php?p...9#pid38799) to create a list of keywords relevant to the RTKO2k1 Lite kanji, and use the kana version of those as keywords? You could probably find multiple words per kanji and select something most similar to the Heisig keyword. Wrightak was concerned that RTK Lite would lose the primitive groupings that Heisig set up, but I think the others already thought of that and made sure to keep them for the most part.

Perhaps you could even head over to this developing thread (http://forum.koohii.com/showthread.php?tid=4308) and get the audio for the words, and use audio cues alongside the kana keywords like how we often do for sentences/vocab in Anki...

From here you can more easily devise strategies and choose when to do sentences for recontextualized reinforcement after you've studied and passed the kanji a few times, or just doing them after given batches of kanji have become fairly mature.

I guess a drawback to this is that you don't want to develop too strong a relation of keyword pronunciation vs. individual kanji reading, hence why I was thinking before it might be better to use English keywords for a while. I think others have already mentioned that though, and wrightak had further thoughts about it.

Edit: Perhaps if you emphasize collocation in the sentences and use pairs of words in kana as keywords, that will make things fuzzy enough to avoid issues with readings? Would also allow for more generalized base to adjust stories around.

nest0r Wrote:while we're waiting for wrightak to stop having a life and come back here with updates...

Not cool wrightak! (Kidding!)

On to serious talk. I think it's really impossible for there to be a "best JRTK", obviously. Somebody early on in their Japanese studies will probably get a lot more out of a "lite" option, but somebody who's far along in their studies it could be completely different.

As somebody who already knows about 1500ish kanji, without ever giving RtK a real try, grouping all the kanji by primitives is most ideal. For me JRtK will probably be a "fast" deck to get through, and hopefully will help maintain my handwriting long term.

I generally don't have problems writing kanji now, but when I do it's always trying to remember which kanji goes with which words, usually on-yomi only kanji. Of course having 10,000+ writing cards would help that, but there must be something less overwhelming...

Anybody else have solutions for this? Perhaps remembering keywords to write the words?

btw I'm Kyeenak from this thread, I made this account because I was giving RtK another try (story reset).

Kyeenak/Grinkers:

*If you don't rely on English keywords, there's no real need to use Japanese prompts which correspond (to the extent possible) with RTK's English keywords. For some kanji, you might prefer to use more common words. Several books give keywords for the kanji (eg Kanji in Context).

*If you already can write kanji without stories, then as you say it's just a matter of maintenance. I don't think that Heisig's primitive order is a good order for adding readings/vocab. You can change the order using Katsuo's lists (stickied under Japanese Language)

*JRTK doesn't really help with testing which kanji is used in words. As you say, there aren't enough words (or they aren't the best choices for that purpose). A compromise might be to use one Japanese word as the prompt (the story trigger) and add additional words (with target kanji in kana) for repeated exposure to other readings and other common compounds. (Now we've come full circle to traditional flashcards ) Books like KinC, 2001KO list common vocab for each kanji. There's also an anki plugin which adds the JLPT vocab for each kanji by level.

*Depending on your level, good ways to practice choosing the correct kanji are dictation (write only the kanji and only words aren't 100% sure of) and keeping a journal.

*I didn't use sentences in my JRTK deck because I didn't want to confuse the sentence content with my RTK story content. (For kanji I don't write often, I still find myself relying on some fragment of a story/image to get the right radical.) To distinguish homonyms, I just added a hint (eg. opposite of __, not___, collocation, example, category, etc). Other solutions might be to include a hidden sentence for difficult words or create a sentence consistent with the story.

Why don't you use Wrightak's current deck as a base and customize it to suit your needs. You could even add vocab to kanji cards as you come across it. It takes time all right, but the process of modifying keywords, etc really solidifies some of the knowledge.

Thanks for the reply, it was an interesting read!

I currently do something very similar to the RtK method, except I use Japanese words (and/or kanji in Japanese words) to build up new kanji. For that reason I think the RtK order would still be useful.

My biggest concerns right now are being able to maintain my kanji writing, because I really don't write that often, thanks to computers. I'm not too worried about remembering which kanji goes to which words when writing rare words. Of course I'd like to be able to do it, but it's not very high on importance with how little is written by hand these days. I figure it's close to me writing in English with a pen, I can't spell at all!

I currently don't do any SRS kanji writing, so I figured wrightak's new deck would be a pretty good starting point. However as far as long term, I'm not sure what I want to do yet. How much work do I want to put into being able to write by hand well (accuracy, special words, and/or speed)?

If someone DOES manage to track the list down and get their hands on it, please post it here for the masses :)

Again, if it is just a matter of data entry, please give me the raw data, and I will make it look nice if need be. I have some time coming up.

I suppose until then I can use the spreadsheet with similar ambitions that was posted earlier in this thread. Maybe some of the words will be the same anyway!

Thanks to everyone who contributed to that spreadsheet, by the way.