It is probably a lot to do with the time investment involved which makes people prefer 6K to 10K. A lot of people opt for the RTK volume 1 Lite version for similar reasons.

That and maybe because the recent updated core 10 is rather new compared to core 6? I don't know.

kill4food wrote:

Thank you, I never looked at it that way. Have you done or are you currently doing Core 10k? If so, do you have any comments on the additional 4k words?

Not going Core 10k at the moment, although I have scoured through the words a while ago and noticed a fair few words in the additional selection of words which I've seen here and there in native media. So I would say that it is probably worth it if you are finishing up core 6 and don't mind that particular method of cramming vocab. I mean, you're going to need to get a lot more than 10 thousand words into your head and familiar, and if the core approach is workable for people then I'd say it's well worth the extra work after core 6.

Thank you, you guys have pretty much confirmed my suspicions. Stopping at 6000 seemed rather early to me and I don't see why you shouldn't continue with 10k unless there is something wrong with the additional 4k words, because even knowing 10k seems very basic to me.

The "Core 10k" data was extracted from the Japanese Sensei Deluxe iPhone app by overture2112 in 2011, and it was originally posted in the japanese 先生 = core 10000? thread. The Core 10k name also comes from that thread. The data is not part of the Japanese Core series published by iKnow / Smart.fm, and there are actually only 9619 pairs of words and sentences. overture2112 posted the data as a Google Docs spreadsheet, but the spreadsheet is now set to private access. The Core10Kv4 (aka core10K (core 2k6k +4k more)) Anki deck includes MP3 versions of the audio files, but I haven't found the original AAC audio files anywhere.

When I compared the data from the spreadsheet uploaded by overture2112 with JSON files downloaded from iknow.jp, they included the exactly same versions of 5544 words and 3739 sentences, but about 1500 more sentences only had small differences.

The Core 6000 and Japanese Sensei Deluxe audio files are spoken by different people, but the Core 6000 audio files are better in my opinion.

The furigana in the Core10Kv4 Anki deck is auto-generated and has a lot of errors. Many versions of the Core 6000 data (like those based on http://forum.koohii.com/viewtopic.php?id=10292 or https://sites.google.com/site/ankinihongo/home/kore) have perfect or almost perfect furigana though.

Last edited by lauri_ranta (2013 December 01, 6:25 pm)

I believe pmnox's 10k deck has 10k unique words, plus jlpt vocab that wasn't in the original 6k data, nor the extra 4k data.  You should take a look at it.

@ktcgx thanks a lot! That's exactly what I need.

@sholum 10k (at least the 10k deck which I found) has audio for both sentences and single vocab words. Also wether 6k words is enough to branch of to mine from native material is debatable and wether it's effective or boring differs per individual I think. Personally I think that 10k words is a better number to branch of.

ktcgx wrote:

I believe pmnox's 10k deck has 10k unique words, plus jlpt vocab that wasn't in the original 6k data, nor the extra 4k data.  You should take a look at it.

It's not true that there are only 1200 new vocabulary items or anything, but pmnox's deck or the original Core 10k data doesn't include exactly 10,000 words, and there are also a few words that are included multiple times.

pmnox's deck currently contains 9888 entries and 9784 unique words based on the literal word column. The non-Core 6000 data in pmnox's deck was based on the Core10Kv4 deck, which contains 9624 entries and 9410 unique words based on the literal word column. The jsensei spreadsheet uploaded by overture2112 contains 9619 entries and 9405 unique words based on the literal word column.

Last edited by lauri_ranta (2013 December 01, 7:07 pm)