« April 2006 | Main | June 2006 »

May 25, 2006

More thoughts on the Heisig method

After writing that last article, I decided to modify my own approach to the Heisig method. On previous attempts I got bogged down while pushing for 100% recall. As the kanji count goes up, so does the review load; going for 100% recall means a huge amount of extra review, leading to slower and slower progress and eventually no progress.

This month (May) I've been trying a new approach. I want to get these kanji learned. I don't have time for 100% recall and hours of review every day. Seeing the kanji again in vol. 2 and in Real Life will also provide repetition and review. So, I've cranked up the pace and am doing about 20 kanji per day. It's not rocket science: it's just memory work. I take 20 or so new kanji per day, more if possible, and really work out memorable images or stories for them. Before bed I review them, and next morning too. Once per week I review the previous weeks' work.

On May 1 I had covered 625 kanji. As of May 25 I've done another 210. That's a pretty nice pace. It's really fun to learn a bunch of new kanji, then look at some Japanese text and find the new kanji. It's a bonus when they have furigana: the readings stick very easily.

It almost seems too good to be true. Since it's a long weekend coming up, I'll do another comprehensive review to see how it's really going.

Permalink | Posted by Joe at 11:28 AM | Comments (1)

May 10, 2006

Some thoughts on the Heisig method

I am a kanji-learning heretic. I have turned away from the One True Church of kanji learning and am following a short, easy path to kanji destruction.

That's what it feels like some days. Last week I participated in a big debate while trying to correct some misconceptions about the Heisig method and really was surprised by the almost religious seriousness with which some people take their kanji learning methods.

They say that the only valid way to learn kanji is by rote memorization of kanji writing, meaning, on readings, kun readings and compounds, all together in one lump. Anything else is not really learning the kanji. If you don't learn them "in context" this way, "you don't learn anything."

A large and growing contingent of Japanese learners is willing to challenge that kanji orthodoxy. I'm not going to give a detailed review of the Heisig method here--I'm working on a separate review of that. If you want to see how it works, you can download a free PDF of the first 125 pages of Heisig's Remembering the Kanji series.

In short, Heisig takes a divide-and-conquer approach: instead of laboriously memorizing numerous different pieces of data for each and every kanji, Heisig advocates dividing up the tasks. First (volume 1) you learn meanings/keywords and writing for 2042 kanji and only then do you learn Japanese readings for those kanji. Rather than memorizing kanji by rote, writing them over and over until they have worn ruts in your brain, you look at the kanji as pictures and stories; the uniqueness of the stories keeps the kanji in your brain. Instead of learning the kanji simply as writing strokes, you learn them each as unique and individual characters with stories of their own.

The Kanji Orthodox object to this division of labor. "You aren't really learning Japanese!" they say, and even "You aren't learning anything at all!" Ridiculous. The Heisig method has a few disadvantages, but overall it's a very effective method for learning.

The main disadvantage of the Heisig method is not that you learn the meaning of the kanji separate from learning the actual Japanese readings. The main disadvantage is that learning all the kanji meanings first means that you delay actually reading the kanji in the context of Japanese text. One of my main goals in learning Japanese is to read, and after almost 2 years of the Heisig method I still can't read Japanese text with kanji. That is frustrating.

However, that isn't the fault of the Heisig method. The blame goes to my own lack of persistence: if I hadn't stopped and restarted twice, I'd have finished by now. That main disadvantage of the Heisig method--delaying learning the Japanese readings of the kanji--can and should actually be an advantage if used the right way. Focusing on learning only the meanings of the kanji means that there's less to learn, and therefore you can learn more kanji more quickly. That means you can learn 2000 meanings in a much shorter time, and therefore can (and should) work harder to do so. Some people claim to have learned all 2000 meanings in a year, a few months, even a few weeks.

However, learning kanji meanings rapidly via the Heisig method has several dangers. The biggest is disbelief: you can learn meanings so quickly and easily that you doubt yourself--you can't really be learning the impossible kanji and doing it so quickly!--slow down, spend more and more time for review, and eventually stop. That's why I stopped the first time: I was aiming for 100% recall and when I didn't get it, I kept reviewing and reviewing until eventually my learning ground to a halt.

The other big danger is to start trying to memorize kanji simply by word and character rather than by associating each kanji with an image or a story. Initially it's a lot quicker to take a pile of cards, write a bunch of characters and think that you're going to learn them by rote repetition, but it doesn't take long at all to start forgetting those characters, lose track of the progress you've really made, and (again) eventually stop.

This time (my third attempt) through volume 1 I'm working hard to stick to the method. Two weeks ago I passed kanji 655, the spot where I stopped last fall, and have been picking up speed. The kanji are sticking and learning them even seems to be getting easier. This morning I hit frame 750 and am more confident in my memory than I was 300 kanji ago.

Another difficulty of the Heisig method is review: after the 500 kanji mark, reviewing begins to take a lot of time. My old method was to review all kanji, repeating and repeating those I missed. I think I'm going to ditch that review method in favor of the must quicker one-shot Kanji Armageddon method. Although I'll lose the multiple repetition of missed kanji, I'm going to be able to move more quickly through the book, and those kanji are going to get their repetition when I hit volume 2 and start adding readings to them.

Kanji Orthodox decree that it's detrimental or a "crutch" to attach English keywords to the kanji first, yet every Orthodox Kanji book I've see does the same thing: you get a kanji, an English keyword and numerous readings and compounds. In the Heisig method, just as with Orthodox methods, the keywords drop away when you've seen the kanji enough times. If the method is a "crutch", it's a crutch you use of a while until you gain enough strength and dexterity to walk on your own two feet rather than a crutch you use because you're permanently crippled. As you learn kanji readings in Heisig's second volume and/or in context of real-life text, the keywords and stories are quickly replaced by real Japanese.

Does the Heisig method work? A growing number of Japanese learners will say "yes."

Permalink | Posted by Joe at 01:43 PM | Comments (0)

May 01, 2006

Kanji Armaggedon Results

Since I haven't done a comprehensive review in a long time, I blasted through all my flashcards this weekend. The review took much less time than I expected, primarily due to the method: each kanji got one trial and either went onto the "knew it" or "didn't know it" pile, instead of my usual method of reviewing less-certain kanji over and over.

Out of 625 kanji, I knew 88% perfectly. I'm pleased with that. Out of the remaining 12%, I was able to draw about half but ended up with misplaced elements. The rest I had absolutely forgotten. I plan to review the incorrect kanji and keep moving: I'm only a third of the way through Heisig's first volume and intend to finish it by the end of the year.

Permalink | Posted by Joe at 02:15 PM | Comments (2)