I said in my last post that Heisig can lead you astray with his ill-chosen English keywords. To give a concrete example, the one I’ve been having trouble remembering recently is 召. Heisig assigns this kanji the keyword SEDUCE, which given the primitives involved (DAGGER and MOUTH) is an evocative image. I never, ever, had trouble with this kaniji before. Then I found out it was wrong. The keyword given by Kanji ABC is SUMMON, which is a very good word to sum up it’s actual usage (召集する: to call together a meeting, or 召し上がる: send for food/drink, for example). [How do I know? I looked it up in a dictionary. But more on that some other time.] For the life of me I cannot come up with a better image Heisig’s. When I see 召 I think “SEDUCE”, and when I review SUMMON I think “…wtf?” because whatever image I came up to associate 召 with SUMMON lost its battle against Heisig’s champion SEDUCE.

As nice as it may be to use Heisig and not encounter these problems (since 9 times out of 10 the keywords, if they differ from Kanji ABC, are easier to remember), it breaks the cardinal rule AIATT: no mistakes. Every kanji for which you learn a misleading or incorrect keyword you will pay for, with interest, later in your studies. I guarantee it.

So what did I do? …I took a page from Heisig. RtK, unlike Kanji ABC, assigns primitive meanings to many characters used as primitives–images which for whatever reason are easier to remember than the character’s keyword, and are only used when the character is used as a primitive in another kanji. Strangely, your mind (if it works anything like my own) rarely has trouble differentiating between the true meaning of a kanji and its primitive/grapheme meaning, as long as you keep that distinction clear from the start.

My modification to the method is the following: if more than once you completely blank on a grapheme, change it’s meaning (this does not apply to kanji keywords). Be sure to check the index to make sure that you’re not using a meaning that will be introduced later. Also go back and change every card in your SRS that references it. This is easiest if you’re using an SRS like Anki that lets you tag cards (which i use for tagging primitive meanings of components). That said, only do this sparingly, when it’s actually required.


Before I say anything, I want to link to a post I previously made on my personal blog on Kanji-learning methods.

I should say is that this was written before I discovered the All Input, All The Time method, and that my comments at that time only make sense in the context of a traditional classroom learning experience (which as we all know Does Not Work). But even so, the flaws I identified in the Heisig method still need to be addressed. To save you a click, here’s my original list of the Heisig method’s flaws, quoted verbatim:

  1. The meaning of kanji is overly simplified. Heisig insists on finding a single English key word to represent the intrinsic meaning of each kanji. While this works for some, it utterly fails for others (for example, 水 can mean not just WATER, but also WEDNESDAY or HYDROGEN);

  2. The form is disassociated from the pronunciation. Heisig lists this as an advantage, but for me it did nothing more than force me to think in English and hinder my ability to later learn the pronunciations; and

  3. No vocabulary or compounds are used. Studying the meaning of a character without regard to its actual use is of little practical value. Especially since the meaning has been simplified down to a single word (see #1) the ambiguity of English has been combined with the native ambiguity of Japanese, making the need for real examples all the more necessary (e.g., does 開 mean OPEN as in to open a door, to open a business, or as in open source software?).

The last two points are automagically solved by the AIATT method–a further testament to its brilliance. My own discovery is that once you’ve fully learned the meaning of a kanji, to the point where once you see it you think immediately of the concept it represents, not the associated English word, there is no problem with confusing the Japanese and English readings (even if that concept you associate it with originally came from its English keyword). And since all focus during the first phase of learning is on the kanji, without simultaneous learning of the spoken language, there’s never a conflict. Brilliant. As to the third point, proper context is of course the what you get when you learn complete sentences, which is the core concept of All (Japanese), All The Time. Trivially solved.

But what about Heisig’s poorly chosen keywords?

Well, that’s still a problem. You see, Heisig developed his method at a point in his life when he knew NO Japanese whatsoever. His sources for keywords were aging Japanese-English dictionaries and the Japanese people around him (a highly unreliable source). Too many times his choice of keyword is a secondary or tertiary connotation, and at times he frustratingly chose a technically incorrect, but related meaning.

The best solution is to avoid Heisig. There is one clearly superior alternative. The book I use is Kanji ABC by Andreas Foerster and Naoko Tamura. Unlike Heisig (at the point in his life when he wrote RtK) these two authors have between them a thorough understanding of the Japanese language and language instruction. The result is a far more accurate book that follows essentially the same (albeit independently created) method. However “Kanji ABC” has its flaws as well–although a description of those flaws and my solutions to them is deserving of a separate post in and of itself.

The key to learning foreign languages (or anything for that matter–I have a rant on that I’m working on) is proper use of an SRS. An SRS, or Spaced Repetition System is an algorithm for managing flashcards in the most efficient manner possible. Current SRS programs use the state of the art in cognitive science to predict how long it will be until you are at risk of forgetting a fact, and schedule your review of the material accordingly. The first review might be in a few hours, the next in a few days, then weeks, months, years… The result is that use of an SRS has enter-and-forget [sic] semantics–once you put a fact into your SRS, you don’t ever have to deal with the logistics of remembering that fact. Just make sure you open up your SRS to review at least once a day and it’ll handle all the details for you

So what SRS do I recommend? For a while I had been using jMemorize, an open source SRS that shows great potential. However at the time of this writing some very critical features are missing, and I can’t say I like the bare-boned scheduling algorithm it uses. It’s open nature put it far ahead of the competition though, and for the last six months or so I had been using my own hacked version of it.

But then I discovered Anki. Anki’s user interface is beautiful. It’s feature set has almost everything you need. And it’s scheduling algorithm is perfect. There are a few things that annoy me. For example, I can’t get cards with images to sync correctly to Anki’s server. But jMemorize lacks support for either images or syncing.. so really, who am I to complain?

你好 and 初めまして!!

I’m jinsei and this blog is a diary of my attempt to simultaneously learn Mandarin and Japanese through the method of constant input, 24/7/365. Michal Wojcik and Tomasz Szynalski were the first to demonstrate the effectiveness of this method when they learned English to a level of near-native fluency in 3 short years. The method was then made popular by Khatzumoto on his website All Japanese, All The Time (and from whom I shamelessly stole the title of this blog).

I’ve been learning Japanese for 3 years now using traditional methods. I’ve been through three different college level instructors, four at-home study programs, and a six week stay in Japan, and through all this I only learned about 100 kanji. And although I was able to ask directions and order food in Japan, and was (usually) understood, I was rarely able to understand the reply I received, and still can’t understand 95% of the Japanese I hear. It was almost enough to make me gave up.

Then I met the love of my life, a Taiwanese student at my university. Not surprisingly, I have now have good reason and a great motivation to learn Mandarin Chinese. However I think she would agree that if I invested the same amount of time into Mandarin that I put into Japanese, and if three years from now my Mandarin was at my current Japanese level, it’d be a complete waste of time. Surely there must be a more efficient way, I thought. And that’s when I found Khatzumoto’s method.

My project is to simultaneously learn Mandarin and Japanese using the method of Wojcik, Szynalski, and Khatzumoto. This blog is where I will document my progress, including rants on what works, what does not, and what changes I make. I consider this an experiment in linguistics: the all-input, all-the-time method has a strong basis in the science of memory acquisition and the psychology of language learning, however very few case studies of its application exist. This blog is my attempt to remedy that, and to assist others who might are trying to apply the all-input method to their own language studies.