Oniichan,

It seems you have two sets which are individually testing the written and audio expression. Is there a reason for this or can they be merged so you test both?

As for the recognition card, that too might be redundant. If you can recall the Hanzi and pronunciation from memory when you see the image + clues, it stands to reason you'll be able to accomplish the reverse.

Nukemarine Wrote:Oniichan,
It seems you have two sets which are individually testing the written and audio expression. Is there a reason for this or can they be merged so you test both?


As for the recognition card, that too might be redundant. If you can recall the Hanzi and pronunciation from memory when you see the image + clues, it stands to reason you'll be able to accomplish the reverse.

Thanks for your help. I think I understand what you are saying. But, my intent wasn't to test the audio at all. The audio was there just for listening/speaking practice. So the templates are designed to test:

image->hanzi (draw it)
image->pinyin (pronunciation)
hanzi->pinyin (reading).

I was thinking of the advice from the old Supermemo list; namely, test one small thing per card.

Do you think it would be better to test audio -> image/pinyin/expression instead or in addition?

BTW, I'm putting together a Core10k for Simplified Hanzi. So far I have my list of words and most of the audio (native recordings).

You have a lot of experience with the Japanese core decks and crowd sourcing etc. What advice would you have for me regarding this project? Adding images is on my to-do list, but what about sentences? It seems that they would be most useful if they are:

1. Natural
2. Provide proper context for the word
3. Increase in complexity as one goes through the list (aiming for something close to i+1)
4. Maybe have audio too.

Are there any good sources for sentences that you know of? With audio? (Tatoeba perhaps)
How about scripts to sort them in i+1 (preferably choosing the best fit when there is more than one available per word)
How should the list be broken up so that this can be accomplished efficiently? (I'm thinking maybe top 500 (in frequency) in i+1 order (but not necessarily frequency order); next 500 (with i+1 based on 1-1000) etc.

Finally, how could the work be shared if there is interest amongst others? So far, nothing I've used as source material is outside the public domain (I think). I'm thinking of something like a spreadsheet (Google docs) with a column for image file names and an online storage folder for the images collected by everyone; a column for sentences; sentence audio; etc. Any advice here?

Or would it just be better to upload to anki and let others edit it there?

nadiatims Wrote:I just really don't get this no translation language puritan approach you occasionally come across online.

Agreed. I put a translation on the bottom of every card I make. Even if I never refer to it, it helps to cement my understanding of the example.

Even beyond that, the ability to switch smoothly between English and Japanese seems almost a requirement for life here in Japan.

JapaneseRuleOf7 Wrote:Agreed. I put a translation on the bottom of every card I make. Even if I never refer to it, it helps to cement my understanding of the example.

Even beyond that, the ability to switch smoothly between English and Japanese seems almost a requirement for life here in Japan.

I'm puzzled as to how learning Japanese via translation could possibly relate to switching between Japanese and English in your daily life.