The Lazy Kanji Kendo Mod | AJATT (2023)

Howdy ho ajatteers! This is tom from Taiwan gonna give you my two cents on The One Method To Rule Them All(tm) for learning characters.

I’ve currently got 2910 active cards in my Kanji deck, which I started making in November 2008, and I can proudly say that I pwn the heck out of chinese characters, which is just to say that I’m speaking from experience. I’ve gone through a number of different formats for kanji cards, including a several month stint with LazyKanji, and I’d like to share with you the results of my tweaking:

When I did Heisig originally I pretty much stuck to what he says in the book religiously, despite running into many of the frustrating problems people are experiencing here. Any tweaks I made in formatting or in my SRS habits didn’t seem to help, and I was always worried that something I would do would make the whole method obsolete, or worse, that “Heisig’s method doesn’t working for me.” Though Heisig and supporters’ talk of his method “cementing characters firmly in your mind” is motivating, please ignore such bold and preachy statements. Heisig’s method is not perfect, nor are you. You will forget characters, you will confuse them. Your memory is a fickle friend, and if there’s one real lesson that Heisig can teach you, it’s not about characters, but about how your mind works.

Anyway, I was doing this originally
Keyword => Character+Story [aka VanillaHeisig]

The problem with this method is:
-Despite being straight from the man himself, I feel that the story here is treated more like an afterthought, where reviewing the story was necessary insofar as you don’t remember how to write the character without it. While you obviously don’t need to memorize the story, per se, I have found that reviewing in this manner is not the best way to use the story.

– As kokage mentioned, also I found it was extremely difficult to recall the keyword when I saw the character in the real word. Intuitively, this is obviously a huge problem, but Heisig dismisses it, and since Heisig was all about going against your intuition I though ok, just do as he says. In other words, I blamed this problem on my own perceived inadequacies rather than accepting that the method was imperfect.

-This has been mentioned before, but when you’re getting into the 3000 range of characters you will definitely need to reuse some keywords, and VanillaHeisig doesn’t accommodate this well.

So there were a few tweaks I used here and there, in card formatting, story creation protocols, using japanese readings, but nothing really took me from feeling like “this isn’t working that well” to the point of “this works like a charm” that everyone else seemed to be feeling. At some point long after I finished Heisig, Khatz mentioned using LazyKanji, and though I had toyed with this idea in my head I didn’t dare go against what Heisig says. At that point I figured, if Khatz is doing it, it must be ok, so I reformatted my cards and gave it a whirl. Lo and behold… still no “ah ha” moment.

I was using this method:
Character => Keyword + story [VanillaLazyKanji]
But didn’t like it because:

-Passivity. I think this is fundamentally the main problem with the Lazy Kanji method, as well as the KendoMod. Yes, its fun and easy like popping bubble wrap but because it doesn’t require much thought, to me it became more like rote memorization than really reinforcing the connections between the character, it’s parts, and the keyword.

-Difficult to grade, since most of the time I’d give a synonym for the keyword rather than the exact keyword, and so the question became “how close is close enough?” Or, from having seen the character in real usage I would know the meaning well, but if this differed from the keyword I just say “oh I know that one” and move on.

-As the cards matured, my ability to remember keywords, or even a synonym, was definitely less than it ought to have been, especially for characters that I would rarely see in regular Japanese. I was forgetting a large number of characters that I once knew, but it was hard to tell whether I had truly forgotten these or if I just had a very fuzzy understanding of their meaning.

With this method, I decided to put the story with a keyword cloze on the front to help with a few of the more difficult cards that I had in my deck, thus making them like the Kendo Mod described above. Then, lightning struck and I came up with a new method, which seems to solve all the above problems in a roundabout way, [Though I have to admit that it is still far from “working like a charm”]

Story [Keyword Cloze] => Character + Keyword

Basically this is the same “KendoMod” that is discussed here with one important tweak: neither the character nor the keyword is displayed in the question. This is a radical departure from what Heisig recommends as well. Nevertheless, I find that it is much easier to both recognize characters and give a meaning when I see them in “real life,” and also easy to recall specific characters when I need to write them.

-Easy/fast/fun to do like LazyKanjiKendoMod, because just reading the story will give you the clues to writing the character properly (i.e. the names of the primitive in the story) while also requiring that you produce the character entirely from memory. I actually find that the small added challenge that makes this more fun and less passive without being too difficult to do.

-Easy to grade, because with a cloze you are less likely to give a synonym. You know easily whether you remembered the keyword properly, as well as whether you wrote the character properly.

-also no problem with having the same keyword for multiple characters

So… to illustrate this I give you a sample card

clothing..display

a […] is a piece of clothing that displays a pattern to wear around your neck

裱

scarf

a scarf is a piece of clothing that displays a pattern to wear around your neck

Notes:
-In the “story” fields, I use bold to indicate the keyword in the story and italics to indicate the primitives.
-Writing the primitives at the top is not really necessary but I find it helpful to type them out when I make the card, especially for ones that are not in Heisig.
-Of course, I also occasionally put pictures on the front.
-After I started learning the readings of characters, I started to add them to the front of the card to help remember the character. I.e. if i was reviewing Kanji and I happened upon one that I knew the reading I would often the card and add the reading. However, to reiterate what I said above, do not bother to do this when you first make the cards. Judge for yourself whether this strategy is helpful or not.

Here’s another one for the road:

hood..one..mouth

[image of the "assassin's creed" guy... wearing a hood with only his mouth showing]

monks walk with their hoods low, they all look the […] and the one part of your body you can see is their mouth

おなじ
どう様（どう・よう）

同

same

monks walk with their hoods low, they all look the same and the one part of your body you can see is their mouth

おなじ
どう様（どう・よう）

Notes:
This is actually an imperfect card because it uses the concept “part of the body” a phrase normally reserved for characters with the 月 primitive. These Japanese readings were added after, AFTER I completed Heisig. After I had them in my sentence cards, and I knew the reading.

Finally, I’ve found there are a few basic characters which are easier to remember just as shapes, or pictographs, than it is to break them down and make a story out of. For these characters I have the same card format, but I don’t have a cloze story. These basically work like this: Story or description => Keyword + Character

For example,

Pictograph of a window with lattices

囪

Window

Normally this would reduce to something like drop+boxed in+wolverine=window, but this too much when its so obviously just a friggin’ pictograph. Zhongwen.com is good at pointing these out.

Thinking about how this relates to the general “principles of flash carding,” keep in mind that what you need to remember is what is on the back of the card: The answer to the question. On the front of the card are the passive pieces of memory that remind you of what you want to actively recall on the back. In my experience, the only way something in the front of the card gets remembered is by sheer exposure: rote memorization. Sure, SRS will give you that exposure, but in order to really strengthen your memory you need reinforce the links between memories as well.

The tricky thing with Kanji is that there are actually two pieces of information that you need to remember for the same character: the keyword and the writing of the character. One way to handle this fact was by having two cards for the same character, one Keyword=>Kanji, one Kanji=>Keyword, although Heisig dismissed this, saying something to the effect if “its a waste of time” and “if you can do it one way you can do it the other way.” For myself, this was partly true, although I wasn’t satisfied with my ability to concretely remember the keywords out of context. Honestly, my current method has not solved this problem completely, though it is an improvement. I would also stress that in practice, needing to recall the exact keyword of a character is rarely necessary. Having a familiarity with a character, being “in the ballpark” or “having a feel for” its meaning is more than good enough 95% of the time. I often feel that I know more clearly what a character *does not* mean more than I know *exactly* what it does mean, yet if I see a new character, I know at an instant if I am not yet familiar with it.

This principle applies to sentence cards too: we aren’t trying to remember what’s on the front of the card i.e. the sentence itself, but we are trying to remember the meaning of the words, the meaning of the sentence, and the pronunciation of the words. Being able to say that sentence or a similar one is less a matter of having seen the sentence several times, but of having remembered the principles behind how its component parts go together.

Coming out the other side of AJATT , the number one piece of advice I could give anyone and everyone here, despite our lust for the speed and efficiency of AJATT– ABSOLUTELY DO NOT RUSH THROUGH HEISIG. Make stories for every character, make them well, make your cards beautiful. You will thank yourself later if you do; you will kick yourself later if you do not.

Your cards do not need to have the Japanese readings on them.
Your cards do not need to have any more complex definitions than the one or two word phrase that Heisig gives. Think of these not as definitions to give you the meaning of the character but are more like a “clue” to get you in the ballpark. You will learn the “real” meaning of them later. Keep in mind that at this stage of AJATT, you are not learning Japanese, you are preparing yourself to learn Japanese. This is a hard reality to take.

I, too, questioned the necessity of this preparation the entire time I was doing Heisig, and it was extremely frustrating. So, I know how y’all feel. When I started Heisig I thought “I would be one of those guys who does Heisig in a month” while in reality, it took me nearly six months. Now, there is no doubt in my mind that learning characters first is absolutely essential, and even when friends, professors, administrators and other manifestations of knowledge-power at my university here want me to study chinese “the right way,” I tell them “No, thank you. I will do things my own way,” knowing that my conscious postponement of speaking will only magnify my pwnage later.

fwiw,
Tommy