+- kanji koohii FORUM (http://forum.koohii.com)
+-- Forum: Learning Japanese (http://forum.koohii.com/forum-4.html)
+--- Forum: Remembering the Kanji (http://forum.koohii.com/forum-7.html)
+--- Thread: A RevTK "Lite" Option (/thread-993.html)
A RevTK "Lite" Option - Nukemarine - 2007-12-06
**Edit - Since it's a long thread, Post 30 and 31 has the list of "Kanji Lite" which are the JLPT2 Kanji and the 92 supporting Kanji. Post 183 on Page 8 combines these into the actual RTK Lite list. There's also a RTK Lite list based off of the 1110 Kanji in 2001KO Books 1 and 2 discussed later in the thread.
Woelpad's RTK Lite Script Kanji.Koohii - Alter Sequence
**
Yeah, this would be WAY down the road but can the following be possible:
In our Profile, there'd a selection option to activate the "Lite" version of RevTK, which will only show you Kanji listed in the JLPT2 list and any Kanji that would be a primitive that's not on the JLPT2 list (such as Axe).
I think the way it could work is this: Each Kanji would have a marker determining if it's in the Lite list (easy enough, well, one would have to make up the list). Now, only these cards will be displayed for studying and reviews. Now, it would mean you'd get a LARGE stack in the added pile but the user would not see it.
So, what happens if he does the 1100 or so that would be the Lite version? I'll assume that the Added stack would be about 900 cards. You don't want to test the 800th card prior to the 10th, so perhaps for added cards, your study screen offers a "learned" button for added cards (much like missed stack). Yeah, you miss out on the initial review, but that's a small loss.
It sounds workable from a programming standpoint (ok, I don't program, but work with me here). Seems like it could make RevTK more approachable. Splits the Heisig system up into more of a 1000, 1000, and 1000 Kanji group. Plus it offers the following:
Student decides he wants to learn Japanese in College, spends 4 weeks in the Summer learning 1000 Kanji at 2 hours a day (very doable). That knowledge gives him a leg up on his first two semesters as they used Genki. Next summer, he decides to round off the next 1000 in about 6 weeks. Maybe then after the 4th semester, he just goes on to RTK3 for the rest.
Thoughts?
A RevTK "Lite" Option - dwhitman - 2007-12-06
My instinct is that this doesn't work, not for programming reasons, but for pedagogic reasons.
Heisig's system is so structured that it just feels difficult to tease out an internally self-supporting subset of kanji/primitives to learn hierarchically, that also meet an external criteria like inclusion on the JLPT 2. Not saying it can't be done, but instinct says "show me".
(My opinion rests in total ignorance of which kanji are actually on JLPT2, just that Heisig's method feels like a delicately balanced house of cards that wouldn't respond well to randomly removing 2/3 of the cards. If it turns out the JLPT2 list strongly overlaps the first third of Heisig, my opinion changes).
A RevTK "Lite" Option - Jarvik7 - 2007-12-06
The entire Heisig method is based upon learning the characters in a way that new characters are built upon previously learned characters and radicals. If you remove a large portion of the characters, the system totally collapses and you might as well just learn using something else. Heisig didn't just throw together a bunch of characters in an order that had nothing to do with frequency or kyouiku grade for no reason.
If a student wants to apply the Heisig method towards learning the characters learned in basic college Japanese courses, he should buckle down during the summer break (first semester classes are just about always in the fall semester only) and get through all of RTK1. It doesn't take that long at all if you're serious about it and use an SRS to help with reviews, as well as stories people have shared here. You can do 100 characters per day easily with the proper study methods.
Alternatively, you can do what I did and just learn the kanji short-term for tests and then cement them into your head whenever you get to them in RTK (I used this method since I had already learned a lot of Japanese before starting RTK). At the very least the short-term memorized kanji should be recognizable when reading, and easily cemented when you hit them in RTK.
A RevTK "Lite" Option - PParisi - 2007-12-06
A good idea for which Heisig's book is not as suitable as "Kanji ABC" or DeRoo's "2001 Kanji." The former work in particular allows you to pick and choose what kanji you learn, as long as you put a little effort into learning the radicals first.
I've mentioned these books in another post so I don't want to give the impression that I'm trying to sell them or convince anyone of their superiority. They are just two other routes to the same goal. Personally, I am squarely in the camp with those who say learning all of the joyo kanji in one push makes sense. Any one of these books would be useful to that end. Because of this excellent site, however, it is much easier to learn and review the kanji using Heisig's text. (Thank you Fabrice!!) Still, there's nothing to stop someone from going out and starting a similar site for those books.
A RevTK "Lite" Option - ファブリス - 2007-12-06
I don't know much about Kanji ABC or De Roo, except for Mary Noguchi's reviews at kanjiclinic... but I don't see why Heisig wouldn't work in any other order. The problem is you need a list telling you what primitives you need. In fact if you were motivated, it's only a matter of jumping back and forth in the book to study only those primitives you need, and go back and study the building block primitives and kanji as required. Sounds cumbersome, but I know for fact someone who did it before, and also had a script to give him the list of primitives to study, depending on the order JLPT, etc. However I'm sorry just spent like 15 minutes searching my mail and I can't find the link anymore, this wa on a student page, it's probably offline now, that was more than a year ago.
A RevTK "Lite" Option - Megaqwerty - 2007-12-06
Jarvik7 Wrote:The entire Heisig method is based upon learning the characters in a way that new characters are built upon previously learned characters and radicals. If you remove a large portion of the characters, the system totally collapses and you might as well just learn using something else.From my understanding, all requisite radicals are still preserved in the RtK Lite scheme, such as axe, even if they themselves are not on the JLPT 2 list. Ergo, the actual structure is still very much preserved, just truncated. The main issue is that quite of the few kanji are not 常用, but rather, just there to better teach radicals (quite a few of the trees serve this function). To that effect, there's no real reason why this wouldn't work, although it might take some time to trim the book: the actual order of the kanji would not change, well, relatively.
But, that's just being a devil's advocate. Seriously, even if you cut the amount of kanji in half, it's better to just let inertia carry you all the way through.
A RevTK "Lite" Option - vosmiura - 2007-12-06
I reccon it should be possible if you selected all the JLPT2 kanji, and all the primitives and kanji-primitives that they use, and also maybe some extra reinforcement kanji just for keywords that don't occur often in just the JLPT2 kanji.
When learning RTK I certainly thought at times "I don't really need all these 60 kanji that use 'thread' primitive". You could certainly prune a lot of them out without damaging the structure.
But I agree with Megaquerty that its probably best to let the inertia carry you to the end. If you tried to build just a JLPT2 book you'd probably still end up with 1300 kanji or more.
A RevTK "Lite" Option - Nukemarine - 2007-12-07
Megaqwerty definately sees the gist of my idea. This RTK Lite would not be arbitrary. It's going in the exact same order. However, if the Kanji is not A) in the JLPT 2 list or B) A Kanji used as a Primitive in the JLPT 2 list; then it is just skipped by the RevTK server.
I think this can be a legitimate advantage of RevTK if it gets implemented later. It's not meant to do me any good, as I did all of RTK (well, still have 400 to get into stack 5). However, that beginner or that college student or that person stuck in a rut can probably benefit.
PLUS, after it's said and done, you unclick the Lite option in your profile and get to go on to finish RTK1 with the benefit of the first 1100 still being in the Leitner mix.
That leaves the questions: Is it effective? Is it doable on RevTK? Will anyone bother to use it? What would be the Lite list?
A RevTK "Lite" Option - vosmiura - 2007-12-07
I was interested in making such a list. I'll look into it.
Its funny now, the JLPT2 kanji list seems so short
.
A RevTK "Lite" Option - revenantkioku - 2007-12-07
It seems like this would be good for only passing a test rather than getting a solid grasp on the kanji. Sounds like a waste, almost.
A RevTK "Lite" Option - Laura - 2007-12-07
I like the idea.
A RevTK "Lite" Option - yukamina - 2007-12-07
revenantkioku Wrote:It seems like this would be good for only passing a test rather than getting a solid grasp on the kanji. Sounds like a waste, almost.Yeah, if you finished the JLPT2 kanji, you'd still be running into lots of unknown joyo kanji not to mention the non joyo...
A RevTK "Lite" Option - vosmiura - 2007-12-07
yukamina Wrote:Yes that's true, but if the book was ordered so that you learned the JLPT2 kanji first and the rest after, then you could finish the JLPT2 ones and start practicing reading for a JLPT2 exam, while still working on the rest of RTK.revenantkioku Wrote:It seems like this would be good for only passing a test rather than getting a solid grasp on the kanji. Sounds like a waste, almost.Yeah, if you finished the JLPT2 kanji, you'd still be running into lots of unknown joyo kanji not to mention the non joyo...
A RevTK "Lite" Option - revenantkioku - 2007-12-07
Guess it's about your priorities? I would like to actually pass the JLPT2 next time I take it (damn 40%!), but passing the test isn't why I'm learning Japanese. Considering it's a year until the next JLPT, anyone starting now has plenty of time to finish RTK1. If you're going to try to speed through kanji for the JLPT anyway, best of luck on any of the other sections.
A RevTK "Lite" Option - Nukemarine - 2007-12-07
revenantkioku Wrote:It seems like this would be good for only passing a test rather than getting a solid grasp on the kanji. Sounds like a waste, almost.No, it would not be a waste of time. It'll be about diminishing returns. The first 1000 account for 70%-80% of newspaper articles. However, it probably accounts for 100% of many instruction materials out there. The next 1000 would then put you upto 95% or so (I'm probably off on that). The next 1000 (RTK 3) account for names and some other Kanji you'll probably run into depending on what you're looking at. Each percentage point you try to cover takes more and more Kanji to get it.
I'm assuming that each 1000 Kanji accounts for about 150 hours of total study time. But that's based off the RTK1 post accounts. I honestly think that most of your JLPT 2 Kanji would be learned at a faster pace based off simplicity (again, an assumption).
Mainly, it is based off my initial post. You can learn 1000 kanji at an easy pace on Summer or Winter breaks. I'm being realistic in thinking most people learning Japanese will be doing it in some sort of college class. I'm self study so all 2000 works for me. For some student with 3 or 4 other classes, maybe 2 hours a day for 3 to 6 months sounds like too hefty an investement.
Besides, perhaps they'll see how easy it is to get that 1000 kanji then go on to learn the entire list fairly quickly.
A RevTK "Lite" Option - Laura - 2007-12-08
Nukemarine,
I think RTK 1 has value, but like some of the critics have pointed out, you'll know if it works for you when you read, for example, a novel, written in Japanese, for educated Japanese adults.
A RevTK "Lite" Option - ファブリス - 2007-12-08
It's tempting, I think something that could work well in the next area (because you can also add vocab etc that uses only those characters).
But for implementing it needs either:
- full list of primitives for each character
- the filtered out list that includes all necessary primitives
I think the second is the best and easiest option. Basically a list filtering out "extra" chararcters outside of the top 1000 most frequent ones (if I understood Nukemarine correctly).
It may be a little confusing though, how will numbering work? The frame numbers need to stay if the student wants to fill the "gaps" later to complete the second 50% missing in RtK1.
A RevTK "Lite" Option - vosmiura - 2007-12-08
I think you need to have a different numbering scheme, with the JLPT2 set coming first and the rest after. Then after you finish JLPT2 you can start adding extra cards at the end of your list.
If you don't change the numbers, when you need to enable the 'uncommon' kanji set, you'd have to set your deck back to zero or else you'd get a thousand unknown kanji start to appear in reviews.
The frame numbers should stay on he kanji display though, so you can look them up in the book.
A RevTK "Lite" Option - Nukemarine - 2007-12-08
Well, I think everything needs to be kept the same as much as possible. You want to keep frame numbers the same in case your stories are referencing them.
Ok, here goes:
1. A seperate chart that tells how many "lite" kanji are in each chapter. It will also say the range of frame numbers for each chapter. Example: Chapter 10, 25 Lite" Kanji, Frames 300-345 (not looking at the book so this is way off). This way you have an idea how many actual cards you will be studying
2. The cards in the added pile have to be treated differently. Instead of a initial test, when you add a card it gets a "reviewed" or "added" button on those cards. When you're in study area, added cards are studied first. My reasoning is that you're adding ranges of cards, but you're skipping A LOT cause you're in "lite" mode. This way, those next 900 Kanji can be systematically studied and added when you enable full mode. In fact, you may need a button to change between studying added cards and studying missed cards (or make it to where clicking on a "learned" button cues up the next missed card while clicking on a "reviewed" button cues up the next added card if any).
Really, I think that would be it. It would be a tad complicated, since RTK was built on the entire Jouyou list. Not sure how many would even bother with it. It feels like a winner.
PS: ファブリス, in the member's list, in addition to showing added cards, could it also show the number of cards in Stack 5 or higher and/or the number of reviews the member has done. The number of added flachcards by itself is kind of useless since many have added the entire list but have done few if any reviews.
A RevTK "Lite" Option - Laura - 2007-12-08
Nukemarine,
I have been trying to accomplish the advantages of what you are proposing by sometimes just leaving rarely used kanji in my orange box. I've also gone through 基本漢字５００ Book 1 recently. I don't think you will ever convince the Heisig detractors that RTK 1 has value, but I do think what you are proposing would allow more people to successfully use the method. Teachers might even recommend it. As you have pointed out, the first 1000 kanji, give or take, are by a vast margin, the most frequently encountered kanji. If you learn these, you have a good base from which to learn words and readings before going back and picking up the other 1000. I think this would balance out the advantages of using Heisig with the disadvantages inherent in depending on English keywords for too long.
BTW I visited yokosuka long long ago with a friend who was connected with the USS Midway. We also went to that weird hotel in Tokyo. Inside it's America: outside Japan.
A RevTK "Lite" Option - woelpad - 2008-01-30
I did a bit of research on this. I copied the JLPT2 list from this site into a spreadsheet, sorted it in Heisig order, then went through the Heisig list to determine which keywords needed to be added. There are possibly some errors in there, so if someone cares to do it over, we can compare.
Here's my list of non-JLPT2 kanji that serve as primitives for the JLPT2 kanji:
吾 冒 呂 旦 白 舌 升 寸 博 貝 頁 凡 句 勺 乙 刀 刃 昭 丁 貫 肖 奇 圭 炎 里 呈 冗 亭 舎 士 吉 茂 是 匕 昆 旨 乞 曽 廷 己 羊 忍 志 串 我 義 戒 刑 乃 及 丈 又 奴 爪 至 丙 瓦 勿 尼 屈 尺 斤 斗 廿 矢 矛 弓 孝 穴 丘 玄 系 卸 厄 宛 酉 豆 凶 亥 契 奉 垂 干 彦 斉 洪 井 亜 氏 司 呉 免
994 kanji in the JLPT2, an extra 92 listed above, for a grand total of 1086. Seems like a nice subset.
A RevTK "Lite" Option - Chadokoro_K - 2008-01-30
woelpad Wrote:I did a bit of research on this. I copied the JLPT2 list from this site into a spreadsheet, sorted it in Heisig order, then went through the Heisig list to determine which keywords needed to be added. There are possibly some errors in there, so if someone cares to do it over, we can compare.Hi woelpad,
I wanted to check to see if there were any errors in your list of non-JLPT2 kanji that serve as primitives for the JLPT2 kanji but first I had to go about copying and arranging the original yamasa list into a spreadsheet.
I don't know how to automate this process so it is taking me some time, but in doing this "by hand" I've noticed some duplicate entries in the yamasa list. (Some are listed under different pronunciation rows but others are just duplicate entries within the same row.)
I'm only about halfway through but so far I've found the following kanji listed twice:
倍 千 弟 捜 札
I'm losing confidence that this is an accurate list of kanji for the JLPT2. (Wikipedia lists the number of official JLPT2 kanji at 1023. Assuming this is accurate, the list at Yamasa is incomplete.)
Does anyone know of/have a more accurate list of JLPT2 kanji?
Also if someone could give me pointers on how to automate the process of dumping kanji into a spreadsheet and moving them all to a single column as well as matching that column with an existing RTK one to get them in RTK order I would be grateful. (For example, I only know how to use "transpose" to change a row into a column one column at a time. I'm also visually comparing the yamasa column with my own RTK column and matching kanji up to get the RTK order number.)
Thanks!
A RevTK "Lite" Option - jmignot - 2008-01-30
I have got a .doc file (downloaded from someWebsite I cannot remember now) called "Kanji in Frequency Learning Order" which contains a table of Heisig's keywords sorted by frequency but with necessary lower-frequency kanji used as primitives introduced at the point where they are needed. Perhaps this could serve the purpose considered in this thread. I can send it to anybody who is interested.
I would also like to be able to leave the least used kanji out from reviewing, even if I have seen them at the time of studying (let them just leak out of my memory !)
A RevTK "Lite" Option - woelpad - 2008-01-30
I'll need to find a better list then. Thanks for the pointer.
I used a slightly advanced text editor, with support for RegExp matching, to create the single column. But even if you're unfamiliar with RE's, this should help. Copy the whole list into a new text file, strip the kana and colons at the start of each line, then replace all spaces with newlines OR (if your text editor doesn't allow that) replace all spaces with commas using find-replace and replace manually all newlines with commas. Save as csv file and import into your favorite spreadsheet program. Your data will now either be a single column or a single row. You can then transpose the row into a column. As you already seem to know, Excel has a special copy feature which on the bottom of the option box has a checkbox for transposing data. Use that to your advantage.
Next up is to get the Heisig list. I used one of the kanji lists from the RTK anki plugin page, opened in the same text editor, replaced spaces with comma's, saved as csv, opened in Excel, selected the first column and copied that to a different sheet in my first csv file. Next to that column, let's say in cell B1, I entered the following formula: =MATCH(A1,Sheet1!A:A,0) and copied that down the whole column. If you then enter in C1 =IF(ISNA(B1),"",A1) and copy that down too, you basically have your original list in Heisig order, albeit with lots of empty cells in between. As a side effect doubles will be blissfully ignored, as such I was not aware of them.
The special copy option to ignore blank cells didn't seem to do much good, but there's still the trusty text editor to remove blank lines, right? Although for the next step, finding extra primitives, you may want to keep the column as is. Alas, that next step is the most taxing, but this should get you started. Now back to finding the JLPT2 list.
A RevTK "Lite" Option - Chadokoro_K - 2008-01-30
woelpad,
Thanks for the pointers and formulas. I now have the yamasa list of JLPT 2 kanji in RTK order. I felt somewhat like a prehistoric person first being shown fire. I don't quite understand how those formulas work but I greatly appreciate the fact that they do. (As a side question where can I look for formulas like these in the future -- or at least the sites that will give me the know-how to write them. Last night I tried to query Excel's help files but couldn't find formulas for what I wanted to do.)
BTW, there are only 982 kanji in yamasa's list. So it is definitely incomplete.
In the forum section of White Rabbit Press (they make and sell Japanese flashcards) the owner states that some sites use outdated JLPT test contents specifications (ie, from 1994) while they (of course) use the latest. This posting was written last year which means that they would have been using the 2004 edition of the JLPT test contents specifications.
(See http://www.whiterabbitpress.com/forum/index.php?showtopic=604&hl=JLPT)
Some where on the WRP forum the owner also mentions that a revision of the JLPT test contents specifications is due out later in 2008.
Wikipedia lists a paper version of the 2004 edition of the book (ISBN 4-89358-281-X) but I wonder if there is an electronic version for sale?
EDIT: The ISBN # actually takes you to the 1994 edition. Several sites talk about the 2004 edition but I can't even find a hard copy of this anywhere let alone an electronic version!