Specifically, I was wondering if anyone has tried memoring Kanji by Bushu (radical) components instead of Heisig primitives, but otherwise using the same method (ie. learning just the meaning and not the readings, and associating each character with one keyword).

I am asking because whilst I like Heisig, I am a bit bothered by the primitives and the keywords not really aligning with the standard 214-component Bushu list. There are some similarities between components and primitives, and also keywords, but there seem to be also differences.

I was thinking perhaps a Heisig like method can be used but based on the Bushu list rather than Heisig primitives. Of course, it requires new stories, but I prefer making up my own stories to Heisig's. And of course, the kanji needs to be memorised in a different order (grouped by Bushu components). A crazy but not totally stupid idea is to actually learn in dictionary order (assuming that you use a dictionary sorted by Bushu radical and in order of increasing strokes) taking care to skip non Jouyou Kanji (if you want). Kind of like some Japanese learning English by reading sequentially through a dictionary.

But the advantage is that there are well known names for each Bushu radical, in Japanese as well as English, which permits creating stories in Japanese if you are so inclined (and fluent enough).

Also, learning by Bushu components will aid searching for words in Bushu-sorted dictionaries and even online (multi radical lookup).

Other advantages I can think of include: grouping the learning process by phonetic components actually allow you to learn groups of characters with similar sounds, which may help the process of remembering the readings later. Alternatively, grouping the learning process by semantic components also allow you to learn groups of related words together.

Disadvantages I can think of: not being able to use this site :-(, potentially confusion by misusing keywords, potentially longer learning process, on my own etc. etc.

Am I being totally crazy? Should I just go back and continue learning the Heisig way or is there any merit in considering an alternate approach?

Yes, I think this is a very valid and sound approach and one that is much more beneficial in the long term.

I completed RTK1 before discovering this site and have done a lot of research and other study since. And I am pleased to say that I have now happily forgotten all the English keywords and made-up names for bushou and phonetic components.

I am well on my way to learning 6,000 Chinese characters now: in doing it I have (loosely) retained the Method but I using the real (i.e, etymologically correct) meanings of character elements. I am also learning the pronunciation and the tones at the same time: any other approach to Chinese would be unreasonable.

Heisig's keywords and names for character components work for someone with a very pragmatic approach to ram the 2,100 characters into memory and move on. For a somewhat deeper understanding of characters (and hence the language) it's better not to rely on his shortcuts.

[ファブリス]

Hi Christine_Tham, and welcome.

Quote:Specifically, I was wondering if anyone has tried memoring Kanji by Bushu (radical) components instead of Heisig primitives, but otherwise using the same method (ie. learning just the meaning and not the readings, and associating each character with one keyword).

Try searching on kanjiclinic.com, I think Mary Noguchi has reviewed several alternatives to Remembering the Kanji, and in particular there is a method from a certain "father De Roo", which works like RtK, but using radicals.

Heisig's primitives correspond for the most part to the bushus, and you could rename the primitives to their bushu equivalent (i.e. "flower" -> "grass") where that suits you. I don't think using Heisig's RevTK system will limit you for using the bushus.

One problem I see with a method like De Roo's, is that you do not get the advantage of the building blocks. With each kanji, you work from the radical components, so with most kanji you have more components to fit into your stories.

Because of that I'm thinking it would be more effective to stick to Heisig's and simply check the radicals alongside RevTK's primitives, rename primitives where you see fit, and learn the remaining bushus that are not covered by Heisig separately.

Quote:Also, learning by Bushu components will aid searching for words in Bushu-sorted dictionaries and even online (multi radical lookup).

Just you wait... I know someone who is working on a very comprehensive and advanced application that will let you search kanji by Heisig's primitives, and I have also implemented a working version of search on Heisig primitives, which is not published on this site yet (it will help finding kanji to add to one's flashcards when the RtK kanji flashcards limit is lifted).

I can tell you by using this search that it is so much easier to find a kanji than using the bushu lookup. Even on the computer the search by original chinese radicals is very slow and complicated. Often times characters are indexed on radicals that don't make sense visually. Have you tried this exetensively? You will be surprised, many times you have to try several differnet sets of bushus to be able to find a particular character.

A search based on Heisig's primitives will be far more efficient because the characters are logically, and visually decomposed; whereas dictionaires seem to stick to the historical tie of the characters, even when the characters have changed form over time and the appearance no longer corresponds to the radicals by which they are indexed.

I don't expect a paper dictionary to feature a Heisig primitive index any time soon, but since people use the computer so much more nowadays I think the a "Heisig search" will be faster for all RtK learners.

Thanks very much for the replies, I really appreciate hearing from people more experienced than me (my Japanese is around JLPT4 from a grammar perspective, but very weak on vocabulary and until very recently knew very little Kanji).

By coincidence, I surfed through the Kanjiclinic articles yesterday and found that Mary specifically advices against memorizing the Kanji using purely Bushu radicals:

To quote Article #51:

Quote:While most radicals--like the 30 examples above--are useful kanji memory aids, memorizing the entire 214-radical set is not a particularly efficient means of learning the shapes of joyo kanji. This is true for several reasons. First, a few radicals are rarely or never used in joyo kanji. Second, in some cases differently shaped radicals have the same or similar meanings (e.g, 鳥 /隹 both mean "bird," 頁/ 貝 "shell," and 皮/革 "leather" ). Third, a surprisingly large number of graphically simple, memory-friendly kanji are not radicals (e.g., 由 reason, 共 together, 才talent, 冊 counter for books ).

Finally, the exclusive use of radicals leads to unreasonable conceptual fragmentation of kanji. Take, for example, the radicals in 輸, the 16-stroke first character in the compound 輸出 (export). Thinking individually about the four radicals on the right-hand side of this character (, 一, 月, and ) makes it difficult to remember the kanji. But if you combine them into the single component , and give it a label (transport), you can streamline the task of memorizing 輸 as well as other joyo kanji containing (愉, 癒, and 諭).

Mary's review of de Roo's "2001 Kanji" book sufficiently piqued my interest that I am going to go out today to see if I can find a copy at Kinokuniya. I may end up buying buying a copy of RTK1 while I'm there (I'm currently learning from a copy borrowed from a library). I have already bought a second hand copy of RTK2 from "Hondarake" (a Japanese second hand "honya" chain that happens to have a branch in Sydney).

I also found (and bought) a book that specifically teaches Kanji through Bushu radicals. It's called "Let's Learn Kanji" and it's published by Kodansha. However, it does not seem to be as easy to learn from as Heisig, because whilst it goes through a lot of trouble describing the radicals (including "non traditional" components) it doesn't provide any stories, or specifically show how the radicals can be used as phonetic or semantic markers (which in my mind would be the key advantage of a radical-based approach).

Quote:One problem I see with a method like De Roo's, is that you do not get the advantage of the building blocks. With each kanji, you work from the radical components, so with most kanji you have more components to fit into your stories.

I don't think (from Mary's description) that de Roo's book is based on Bushu only, I understand his approach uses component/primitives similar to Heisig, but has a different set. However, he doesn't specifically impose a character learning order, so you can more or less learn the characters in any order (he repeats the meaning for components in each character they appear in). This is attractive to me as it allows me to use stories to remember kanji sorted by school grades, JLPT etc.

Quote:Because of that I'm thinking it would be more effective to stick to Heisig's and simply check the radicals alongside RevTK's primitives, rename primitives where you see fit, and learn the remaining bushus that are not covered by Heisig separately.

I think you are right. For the most part, Heisig's keywords for primitives that correspond to Bushu are similar, with a few exceptions, such as the "flower" vs "grass" example you mentioned. I may try and substitute Bushu keywords for Heisig for my stories as you suggest (where they make sense, and particular if it makes a better story!).

Quote:Just you wait... I know someone who is working on a very comprehensive and advanced application that will let you search kanji by Heisig's primitives, and I have also implemented a working version of search on Heisig primitives, which is not published on this site yet (it will help finding kanji to add to one's flashcards when the RtK kanji flashcards limit is lifted).

Are you talking about something similar to the multi-radical lookup implemented in WWWJDIC and programs based on KANJIDIC (eg. JWPce)? If so, I agree, it's much more useful than traditional Bushu single primary radical lookup - it's my favourite way of finding a character (even more useful than SKIP). Using Heisig primitives would be a good alternative though for RTK graduates!

I will be very interested in being a "beta tester" for your friend - I am always on the lookout for good computer based study aids. I use JWPce and JFC all the time as I have them loaded on my mobile phone (running Windows Mobile).

Quick update - went to Kinokuniya, very disappointed. They must have had a sale recently, the Japanese language textbook/dictionary section seems somewhat depleted. Not only could I not find de Roo (not surprising) but I couldn't find Heisig RTK1 (they had 2 copies of RTK2).

Then I went to the Japan Foundation library and they had all the major books recommended by Mary in Kanjiclinic: de Roo, Henshall, Foerster/Tamura (Kanji ABC). So I borrowed all of them (have now maxed out my borrowing privileges so can't borrow anymore).

de Roo is interesting, not what I expected. It's basically a mini dictionary, focusing only on the Jouyou characters, based on a unique indexing system (identifying the shape of the top left and bottom right parts of the character, then generating a pair of 2 digit numbers corresponding to 2 sets of shape lookup tables). It gives a small subset (the most common) on yomi and kun yomi readings, and a very concise story to help you remember the meaning (sometimes two or more stories are given per character, corresponding to the different readings) linked to the "primitives" (or "graphemes") as de Roo calls them.

de Roo has a listing of the graphemes up front, these are numbered quite strangely (3a to 39a). They appear to be based on a subset of the Bushu radicals together with common components that are not traditionally classified as radicals. Each grapheme is associated with one or more keywords, and a cross reference (presumably non-exhaustive) of kanji words that use it.

The main book contains the Jouyou list sorted by de Roo's indexing system - from (3,40) to (39,79). What's unique is that each character is cross referenced against a set of other characters related in situation, meaning or shape.

I like de Roo's stories - they are very concise and to the point - much better than Heisig's tendency for bloat, obscure/obtuse references and sometimes meandering. The book is worth buying just for the stories.

The print quality is not great - it's essentially a set of hand drawn characters (neat calligraphy though) and type-written text, plus romaji for the readings.

I'll give my impressions of Henshall and Kanji ABC in the next two posts.

Henshall's "A Guide to Remembering Japanese Characters" is also based on components (Henshall calls them "elements"). Unfortunately these are only loosely based on Bushu (much like Heisig primitives).

Heisig goes through the characters in grade school order. He provides on- and kun- readings (romaji), stroke count, and a representative set of compounds. He also includes a paragraph describing the component breakdown into elements, and a story which often explains or suggests why these elements were used and how they relate to the overall meaning. Finally, Henshall provides a 1-line mini story or "mnemonic" to help remember the characters.

Henshall's mnemonic is not as good as de Roo's, and can sometimes be rather obscure if you haven't read his paragraph on the character. But he wins marks for being concise.

I think Henshall is a good supplement to Heisig - often his paragraphs explains Heisig's stories better than Heisig, or provide a plausible reason why the element combination is so bizarre. An example: "employee" in Heisig is "mouth" over "shellfish" which is bizarre but Henshall explains that the shellfish component is originally a different component.

I wouldn't use Henshall as the main method for remembering Kanji though - I think Heisig is a simpler and easier. Rather it's a reference text when Heisig's explanation stumps you.

Final mini review: Foerster and Tamura's "Kanji ABC".

This book is written after the others (in the early 90s) and benefits from lessons learnt from approaches taken by the other books. Essentially it is a refined version of Heisig, but fixes the issues I mentioned in my original post.

First of all, the graphemes are a proper superset of the 214 Bushu radicals (yeay!) - missing a few rare ones not used in the Jouyou set.

The book has a much better ordering of characters compared to Heisig's somewhat haphazard sequence. Graphemes are grouped into sets, and kanji characters are also grouped into the same sets. Sets tend to connected by common reading or meaning, hence addressing my point about taking advantage of phonetic/semantic markers.

Each character is provided with on- and kun- readings (romaji), keyword, grapheme decomposition, and cross references to Nelson, Halpern and Spahn dictionaries.

This is a book that contains the bare bones information to make up one's own stories. The authors themselves haven't provided any stories at all. So it seems primarily aimed at someone who likes the Heisig method but wanted to make up their own stories and addresses all the points I made in my original post. In other words, this book is targetted for someone exactly like ... me!

Conclusion: I haven't decided yet, most likely I need a combination of all books - I like the systematic approach of Kanji ABC, the stories from de Roo, and the background context from Henshall. And of course, the main advantage of Heisig is this web site and community that has developed around it. Tough choices ahead.

[ファブリス]

I would recommend you to stick to Heisig and use the excellent Kodansha's Kanji Learner Dictionary to fill in all the background information you need about the characters. This is a really good combination, as KKLD also features keywords for each character.

Perhaps your main concern with RtK is the strangeness of the stories. You would not be the first person to be put off by the "imaginative mnemonics". However this is just a system, and imagination is a very powerful tool for remembering. It's just a fact that memory links become stronger when emotional response are attached to it, so weird stories and mental images often work better than logical explanations; except for those few simple kanji where the primitives fit so well with the keyword that it's hardly necessary to create a mental image for it. Keep in mind that Heisig only provides the first 512 stories, and you are free to remake all of them. But you have a solid system already laid out.

When you say Henshall wins marks for being concise, it all depends on what kind of memory you want to use. Some people can remember very well with the use of very short mnemonics which hold in memory by the arrangement of words and maybe their sounds. Most of us can work with such mnemonics and have used them without really thinking of it; but I think most people can not use them reliably for over 2000 kanji. Concise stories/mnemonics will otherwise be not so helpful, like Heisig says in RevTK1, he begins with detailed stories, to help you flesh out the most vivid and detailed mental images, that will hold on better in memory. He then moves on to more concise stories near the end of the first Chapter and invites you to detail them with your own memory associations. But the point really is that you need a good mental image. For me and many learners, it's possible to do that with eyes closed or open and just build up things in the mind, but for many people this is more difficult at first, I guess it's a left brain /right brain thing, so for many people detailing a story, even typing it in on this website or writing it down, is another helpful way of conjuring up these mental images.

PS: thanks for your detailed comments on DeRoo and Henshall I will rename the topic accordingly for the forum search.

ファブリス Wrote:I would recommend you to stick to Heisig and use the excellent Kodansha's Kanji Learner Dictionary to fill in all the background information you need about the characters. This is a really good combination, as KKLD also features keywords for each character.

I agree with ファブリス. This is how I modified the Heisig system for myself. I also used it and "Kanji in Context" to learn vocab at the same time. (But I already had a strong base in spoken Japanese. If you don't yet, then it is probably best to use KKLD only to modify the keywords as necessary at this stage.) Be warned, this is more work than just going with the Heisig system as is.

ファブリス Wrote:When you say Henshall wins marks for being concise, it all depends on what kind of memory you want to use. Some people can remember very well with the use of very short mnemonics which hold in memory by the arrangement of words and maybe their sounds. Most of us can work with such mnemonics and have used them without really thinking of it; but I think most people can not use them reliably for over 2000 kanji. Concise stories/mnemonics will otherwise be not so helpful, like Heisig says in RevTK1, he begins with detailed stories, to help you flesh out the most vivid and detailed mental images, that will hold on better in memory. He then moves on to more concise stories near the end of the first Chapter and invites you to detail them with your own memory associations. But the point really is that you need a good mental image.

I ended up giving away my Henshall book because his concise mnemonics were useless for me--they didn't stay with me at all, and that was only at a few hundred kanji. Now this might not be the case for all other people, but I'm fairly sure it will be true of a lot of people. Concise is not necessarily better, it is only better if it is still memorable for you. And there are some very memorable and concise stories offered on this website.

If you decide to go with Henshall or Kanji A-B-C, do let us know how well they work for you, why they work for you, how you modify their system, etc.

Good luck in your studies!

ファブリス Wrote:When you say Henshall wins marks for being concise, it all depends on what kind of memory you want to use.

My apologies for giving a misleading impression of Henshall.

He doesn't just give concise mnemonics (otherwise it will be a very short book). He actually gives fairly lengthy stories with each character, in many cases rather long explanations. However, he then summarises the story in a concise mnemonic, which I found very helpful.

A lot of Henshall stories relate to the etymology of the character.

I don't think Henshall will be a good book to learn from - my eyes tend to glaze over some of the etymological explanations. But I think it's a useful companion to Heisig - if someone is struggling with remembering a specific character, then Henshall may provide just the right background to lock the character in memory.

I still haven't decided yet which is best, but I did end up ordering all the books I borrowed from Amazon. So I'll probably end up using a combination of all four - I like this site and probably will still use it, but I'll supplement my learning with the stories from Henshall and de Roo. Kanji ABC on the surface seems to be everything I want, but the lack of tools and community support means I'll be pretty much on my own and that's a scary prospect.

Thanks for the advice on the Kanji Leaner's Dictionary. I bought it a while ago, but haven't really used it (mainly because I find JWPce lookups much more convenient). Might give it another go.

The main point I gathered from people's comments is to build as much context around each character as possible so that it becomes "alive". After all, our long term memory depends on how many associations and linkages are built. So I think the approach of using multiple texts should help in this regard.

hi there i just skimmed through the conversation but just wanted to add that basically if you have 5 hours a day you could complete the entire RTK in 8 weeks with some days for review only for when it all piles up. Thats what i did. The time IS there, you just need to find it. I spent 30 mins while eating brekkie reviewing, 30 mins on the bus making new stories, 1 hour in my lunch break with anohter hour behind bosses back doing a mix of review and new stories, 30 mins on bus on the way back formulating stories and then the rest at home to do what was left to be done and still had a social life.

I think its addictive to try and look for an even better approach ( i do it all the time, new SRS, new PDA, new this new that) but if you spend 4 weeeks researching, finding and buying the material you could have already learnt 1000 kanji the RTK way.

I know you were talking in respect of the overall learning process and im not at a stage to comment if hesig is overall quicker than other systems however i have started to use the approach outlined by japaneseallthetime.com (sorry its not the exact url but has been discussed a lot on these boards so do a search) and have found retention of the pronunciations a lot easier than expected. And the guy behind that site has got an amazing grasp of japanese in a very short time (1 year and 6 months i believe). Basically after learning all the kanji in RTK you go onto total input of films, books, music etc. From constant reading and reviewing of new japanese ive found tackling kanji quite easy.

I reckon you should give yourself a week to find a better newer method but if it doesnt turn up just go the heisig way and really commit yourself to it and in 2 months youll be on the "self congratulation board.". Though pioneering some new approach is appealing how much more time can you shave off the rtk method or the japanese all the time guy?

Ganbatte!

Quote:if you have 5 hours a day you could complete the entire RTK in 8 weeks with some days for review only for when it all piles up.

I agree - if I really had 5 hours a day that's what I would do. Learning 200 kanji a week will essentially allow me to go through RTK1 in 10 weeks, and I kind of proved that I could do that (in 3 days I managed to chalk up 172).

Unfortunately, I can't really sustain that kind of pace. Realistically I would be lucky if I can afford 1 hour a day, and at that rate it will take me a year to go through RTK1.

And I discovered that just associating meaning with the symbols is not very helpful. I don't really understand the nuances when the character is used as a verb vs a noun, and how the meaning changes in compound words.

Last night, I spent about 2 hours trialling a different method (of course, I made sure I reviewed all the expired kanji on this site first - boy I have already forgotten a lot!).

What I am trialling is basically reviewing the JLPT4 kanji by the following process:
- for each kanji in the JLPT4 list, look it up on JWPce using the kun-yomi reading or using multi-radical lookup
- write down all the relevant indices (#strokes, Bushu, Heisig index, Halpern index, De Roo index, Henshall index)
- now look up the character in the Kanji Learner's Dictionary and practice writing the character using the stroke order diagrams. Try and get a sense of the multiple nuances of the character in the dictionary - different forms, how the character changes meaning etc. At this stage make no attempt to memorise any of this, it's to get a "feel" for the usage of the character.
- Look up the character in Henshall, read the story and copy the mnemonic.
- Look up the character in De Roo and copy the story. At this stage I am not bothering to look up the De Roo cross references.

Man o man it's tough going. It takes me around 10 minutes to cover each character so in two hours of solid studying I only managed to cover about 16 characters.

But the advantage is I really feel I know each character quite well after that. And the bonus is I am also learning the graphemes as well - for example, ta(beru) (eat) is essentially a roof over the grapheme for "good", which is also an independent kanji. From Henshall and De Roo, I also learn how the character originally represented eating white rice using multiple cups under a roof, which is a very powerful story for remembering how to write the character. Henshall is great in that he often debunks the usual explanation for a character that they teach in Japanese schools, which I found interesting (I'm going to try and remember some of these and use them in conversations with other Japanese students).

This seems a good way to allow me to learn how to write the JLPT4 set (previously, I could only recognise them sometimes and didn't know how to write them). I'm hoping to go through the JLPT3 and 4 set in a month or so (if I can struggle to achieve 20 characters a day I can cover JLPT3 in 15 days, but realistically I will be lucky to achieve 10 a day).

I'm still planning to go through RTK1 as well via this site, but more for fun (I like making up stories, and they help me with the other process as well). My other method should allow me to aim for the JLPT3 test this year if I am so inclined (I'm not, I hate exams and have no wish for a paper accreditation). The real advantage of knowing the JLPT3 set is being able to breeze through textbooks without glancing at the furigana, and just focuse on the grammar and usage.

Okay, I've covered about 36 characters now using my "slow" method" (I was hoping to do more, but this weekend was rather busy).

Although very slow, the method does seem to be working. I seem to have no problems remembering kanji that I've previously struggled with, and indeed the characters now conjure up a rich tapestry of reading, meaning, etymology that they didn't before.

I do think that learning the characters in JLPT order is the right thing to do. First of all, I already know the readings and meanings for these characters, so I can focus on learning how to recognise and write them - in other words, I am approaching them from the perspective of a Japanese child. Secondly, these characters are also roughly in school grade order, allowing me to practice the readings using the myriad of kanji practice books that one can buy cheaply (I picked up a stack of Grade 1 to Grade 6 "kanji practice books" for A$2.50 each at our local Japanese supermarket).

I must say, I have grown to respect De Roo enormously. His book may be thin, badly printed and very hard to find, but it is an absolute treasure. He seems to have summarised basically a lifetime of studying Kanji into 1-2 lines per character. I tend to read De Roo after reading Henshall, and time and again I am amazed at how De Roo summarises the gist of what Henshall says in a long paragraph into a short sentence. He somehow encapsulates meaning, component analysis and etymology (plus hints on how to decode the character as a grapheme in other characters) in no more than 10-20 words. Amazing.

The De Roo cross references are worth checking out. I went down 1-2 chains and had a few "a ha!" moments.

I have now fallen behind in advancing using the RTK1. I'll try and do one or two lessons sometime this week as well.

This may have been stated, but if you learn by bushu, you won't be getting all the components that make up the kanji(a great point about RTK). Before I started RTK, I tried learning all the bushu, but some of them were pretty complicated and useless, and of course, I'd come upon kanji that had non-bushu components...

I often don't get enough of a feel for the kanji meaning just form one keyword. I like looking them up in JWPce to see what the other meanings are pronunciations are, and what words use them. Reading the etomologies in Henshall can really help sometimes if I'm having trouble too.

I find JWPce the best way to look up kanji. I usually use the multi-radical look up, but you can search by keyword(Heisig keywords will always work), reading, stroke count, etc. Much faster than a book and not slowed down like a website might be. I can almost always find what I'm looking for.

yukamina Wrote:This may have been stated, but if you learn by bushu, you won't be getting all the components that make up the kanji

Yes, it has already been discussed, see for example my posting in Reply #4.

However, it *is* possible to pick a set of primitives (or "components" or "graphemes" or whatever you want to call them) that are a proper superset of Bushu, and there are a number of advantages of doing so.

See my mini-review of "Kanji ABC" which takes this approach - their graphemes are much better organised and intuitive compared to Heisig.

The bushu are not as complicated or useless as you may think. For example, "Let's Learn Kanji" has a very good breakdown of the more useful bushu and explanations of how they influence the character from both a phonetic and semantic perspective.

De Roo also provides some excellent clues through his cross references on understanding the interaction between bushus and the characters.

ファブリス Wrote:I have also implemented a working version of search on Heisig primitives, which is not published on this site yet (it will help finding kanji to add to one's flashcards when the RtK kanji flashcards limit is lifted).

I was recently thinking that a primitive search (useful for when you remember the wrong keyword for a kanji when reviewing but then are unable to lookup the correct one in order to note it there too) could be another application of the hypothetical primitives list (http://forum.koohii.com/showthread.php?tid=611 http://forum.koohii.com/showthread.php?tid=221) and that you could generate one by processing the story markup on this site. How did you make your list, and are you able to share it?

[ファブリス]

cangy A list of distinct primitve elements as discussed in those two topics you mentioned is in Index II at the end of RtK Volume 1. It's not very long and if it was really important to someone they would have produced the list already. The workaround for the font that was suggested was to use kanji that include the primitive.

You are talking about another kind of list though, one where each kanji is mapped to its constituent primitives. To my knowledge there is no such list readily available. However there is something called radkfile. I used this to get started. I manually entered some of Heisig's primitives, and simply renamed others. It's not 100% accurate.

Your idea to process the story markup is really good! Never thought about it. That said, for my purposes I would need to be able to map the primitives to thousands more characters outside of RtK1 and RtK3. I currently have a textbox where you have auto-complete like in the Study search box, except it works multiple times for entering primitives. Using this, I could setup a small app that lets users index all the kanji. That's what I had in mind. But with the modified radkfile I'm using the results are helpful enough that it could be released as such before the correct indexing data is obtained. I've used it quite a few times now, it's far easier to type in two or three primitives with an auto-complete dropdown than to search in a grid of chinese radicals where you are also forced to mentally count everytime you need a radical.

I have heard a lot of good stuff about DeRoo, and I'm really enthusiastic about finding a copy of his text, but it seems to be hard to come by. Does anyone know where I can order a copy from?

Also has anyone else used it? What were your impressions?

I don't know if Christine Tham still visits this site, but I'd like to thank her for writing such comprehensive reviews of Kanji learning materials. I know most people on this site use Heisig, but it is good to see a variety of approaches discussed !

Personally I'm using Heisigs method but I still know at least 50 bushu. I seriously find no problem in using both, I learn using Heisigs method but I learn the bushus anyway, both stick. I mean, Heisig is just a story to let you write the kanji, what he calls the primitives and what he calls primitives isn't really important in the long run. For example, I know the radical in 持 is a hand even though Heisig calls it a finger, that really doesn't hinder me in any way.

Thanks for your comment Tobberoth ! It seems the text by DeRoo is definitely out of print, so I might have to find a different book to study from or hope it is reprinted in the near future.

With modern technology I am surprised books go "out of print", surely for a price a publisher could just print a copy on demand.

MeNoSavvy Wrote:I have heard a lot of good stuff about DeRoo, and I'm really enthusiastic about finding a copy of his text, but it seems to be hard to come by. Does anyone know where I can order a copy from?

Also has anyone else used it? What were your impressions?

I don't know if Christine Tham still visits this site, but I'd like to thank her for writing such comprehensive reviews of Kanji learning materials. I know most people on this site use Heisig, but it is good to see a variety of approaches discussed !

I saw copies of both Kanji ABC and the DeRoo book in the NYC Kinokuniya bookstore.

Thanks for the information playadom. I checked on the Kinokuniya website, but it doesn't seem to be mentioned on their online catalog. It may just be that the NY store had a couple of copies. Or perhaps they have sold out since you were there?

If anyone comes across an online store that has copies of deRoo please let me know. Thanks !

I notice there is a used copy on Amazon at the moment for $100 ! Yikes !!! With the current exchange rate that is getting close to $200 NZ dollars, ouch ! Without looking at the book first I'm not sure I'm willing to pay that.

I don't actually live in NYC, but I get around to going to the bookstore every few months. The books were pretty tucked away, so it doesn't seem likely that they would be sold out[although this was at the beginning of September].