Ok, just had an idea and am looking for opinions

what would be the effectiveness of taking a large portion of vocab 10kish in this manner, assuming you had a basic grasp on conjugation and sentence formation.

1. hear/read word e.g. saifu
2. read definition e.g. wallet
3. read part of speech e.g. noun
4. hear/read word in a sentence e.g. watashi no saifu nai
5. form a mnemonic quickly e.g a man with a camera saying sai fu! then at the flash of the camera grabbing your wallet and running
6. next word

this taking about 5-10 seconds per word

based on the effectiveness of mnemonics
and the fact that 8000-9000 word families make up 98% of vocab in most conversations, news articles and novels.

would this effectively add that word to your passive vocabulary?
any sugguestions to of how to add or improve on this method?

semperanimus Wrote:would this effectively add that word to your passive vocabulary?

Why don't you try it yourself and tell us the results?

I will.
trying to find an appropriate list atm.

haha. Just 2 quick thoughts before you go off experimenting: :-)

-You might not need 3 (pos) as it will be apparently from the sentence.
-I'd suggest saving word mnemonics for the occasional word that's not sticking or when it's almost unavoidable from the meaning of the kanji. Even then, it'd be a very temporary crutch. I think you'll get bogged down attempting to create one for every word.

about 3, I think it would be useful for na and i adjectives, but your right about it not being important for most words.

the point of mnemonics is so that you won't be bogged down in memorizing. and so that you can make words stick easy. remember we are going for rapid memorization here.

thanks for the sugguestions though

I don't think the part of speech is necessary in terms of making it a deliberate study element, so much as something you kind of glance at while looking at the usage of the word.

For the mnemonic, if you're gong to use the yomi in it, make it meaningful somehow, like wordplay, like perhaps in this example it's a warped version of that baseball thing where they yell “Safe!” (as they're getting away with your wallet). (The cool thing is that as you learn more vocabulary, you can use other Japanese words to inform this wordplay. It's not a big deal to use English stuff since they're transient structures that often are forgotten by the time you next review the card, but it's a great indicator and good practice when you have enough readings internalized that you can apply them to new words.)

I'm glad to see someone else who appreciates that mnemonics don't have to be dated, tedious structures (which are still more effective than rote) but instead can be quick encoding tricks that spare you tonnes of time and effort and make long-term retention much more effective and easy.

thanks nest0r, great idea about building mnemonics using Japanese words already covered, that could be a helpful way to review that wouldn't slow you down.

semperanimus Wrote:I will.
trying to find an appropriate list atm.

I recommend the shared deck called JapaneseCorePlus. Over 20,000 words, ordered by frequency. Sounds exactly like what you're looking for.

kinda offtopic
thanks vileru, I downloaded the deck but I'm having troubles.


after download I clicked start review and I got through the first few cards before I realized there's no audio! I think the description said it had audio, maybe I have to change settings or something? how do I make the front side say the word and the backside say the sentence aloud? I tried searching anki faq but apparently this isn't a frequently asked question.
I use windows XP btw.

And I need audio if I am to get correct pitch accent for each word. (I think)

I think your estimate of 5-10 seconds per word is extremely optimistic.

Learning 10kish words is a very good idea but I would skip 3, 4 and 5.
3 because pos becomes obvious to you when your grammar level picks up, so you don't need to memorize it on a per word basis.
4 because if you're adding vocab from reading material or phrasebooks or whatever source you're using you already have seen a sample sentence anyway. If you're taking vocab from frequency lists or something, then searching for and selecting example sentences is gonna take time.
5 How do you know if the mnemonic is strictly necessary if you haven't tried learning the word without it. Learn a word first without a mnemonic then if and only if you persistently can't recall it on later reviews then spend the time coming up with a mnemonic.

Using anki, search for a shared deck called 'Japanese Core 10,000 with audio v2' and also download the audio files as per the uploader's instructions.

The method you are saying to learn words remind me to the book "Aprende un idioma en 7 dias" (Learn a language in 7 days) by Ramon Campayo. It basically presents the idea of learning the most frequent words in a language (about 1000 I think, I dont remember well). He puts the word in the langugage you wanna learn, the world in your own language and the mnemonic. He separates the words in the different parts of speach (nouns, verbs, adjectvie).

Here is a public spreadsheet with the talbes (althought, the translations are in spanish) https://spreadsheets.google.com/pub?key=...yUlE&gid=0

So, well, I haven't tried it actually, but it seems like a good idea I was gonna learned the words this way, until I discovered the core2k shared deck and loved it 'cause the images, example sentences and sounds. So, decided to apply a similar method, but using the words in the core.

For the nature of Japanese the part of speech is needed. 冷蔵（れいぞう） is both a Noun and a verb depending on it usage. The sentence example should be simple, but grammatically correct. Remember that learners of various level might find your tool very useful. Hint could be useful for beginners.

When 冷蔵 is used as a verb, it's going to have a verb like する appended to it which will prevent any confusion. Memorising part of speech is a pointless waste of time. Just gradually get a sense for how a word is used by reencountering it in different contexts.

The fact is that 冷蔵 is both 名詞 and 他動詞, as that is how it is listed in Japanese dictionaries (eg. 明鏡国語辞典 and 大辞線). One of the things that is intriguing about Japanese. Maybe several sentence examples would be beneficial to understand the nuances of Japanese.

this is assuming a pre-made deck with words you've never seen. (so you haven't seen the word in a sentence) and also assuming the mnemonics are just 4-5 second things for you like they are for me.

semperanimus Wrote:this is assuming a pre-made deck with words you've never seen. (so you haven't seen the word in a sentence) and also assuming the mnemonics are just 4-5 second things for you like they are for me.

Pfft, n00b. It takes me 2 seconds! The meaningful iconic nature of kanji is especially help here, as you learn words/have finished RTK. In that sense, the relational strategies/encoding tricks are built in, so often it comes naturally without custom effort. It's just a matter of making the relations explicit to yourself when studying new/failed cards.

semperanimus Wrote:this is assuming a pre-made deck with words you've never seen. (so you haven't seen the word in a sentence) and also assuming the mnemonics are just 4-5 second things for you like they are for me.

4-5 seconds seems very optimistic to me, especially after you've studied more than 30 minutes or so and you're approaching break time. I think in practical terms it's going to take you more than 5-10 seconds to go through your 5 step process. That's just 1-2 seconds per step. How is a learner supposed to read a japanese sentence which will include unknown vocabulary that quickly? Some words will just be a bitch to make a mnemonic for. Try making a mnemonic for words like そっくり, きっちり, etc. Realistically, I think you'll burn yourself out trying to go through all those steps in that amount of time...

Took me a total of 3 seconds to come up with mnemonics for both そっくり and きっちり. This, multisensory integration, and SRS using my method for learning new words just hypothetically saved me 90% of the effort and time it would've taken nadiatims to learn those words or some equivalent words in the long term. I was then able to learn many more words in less time, learning them for longer, and able to reinforce them in native materials with 90% more hypothetical enjoyment.

uhuh...You went and timed it I guess...? What are your mnemonics btw, assuming you still remember them?

Of course I timed it, using an atomic watch. I actually exaggerated the time it took out of modesty. I could tell you the mnemonics I came up with, but they're so brilliant I'm only selling them. 500 dollars. I guarantee a refund but only after 18 months, to weed out uncommitted people. Suffice to say they were highly effective relational strategies that improved encoding by 90% and reduced the time I spent reviewing them by 90%, giving me 200% more time to learn 1000 other words and read/hear them all in native media in the same time period you spent learning the past 10 words you learned.

[Katsuo]

I don't know what nest0r came up with, but using a standard "sounds-a-bit-like-X" mnemonic method:

そっくり (just like) sounds a bit like "socks", so imagine pairing socks after doing the laundry, i.e. picking ones that are similar (そっくり)
きっちり (fit tightly, exactly) sounds a bit like "kitchen", so think of a fitted kitchen where the cooker etc. all fit snugly.

But does that stick better than seeing/hearing そっくり in this advert?

not bad not bad.

But you already knew those words
Also they took you more than 4-5 seconds each right?

I have a slightly tangential question.

When people refer to the "X000 vocab will cover 95% of convos"...

Does this statistic account for different forms of verbs?

行く
行きます
行って
etc etc

The word families stuff, for English, refers to multiple word forms. 8000-9000 word families = ~35000 words. As I mentioned elsewhere, as that percentage climbs, the number of word families in the last ~3% is much bigger, like thousands of word families. That's a summary of the breakdown in N. Schmidt's paper I posted before (you can search for it here if you're interested) but am too lazy to link again.