Fluent Forever

The new home of Tower of Babelfish

Modifications for Japanese and Chinese

This page is here to show you how to modify my methods for Chinese and Japanese.

Both of those languages use logograms (characters that represent meaning, rather than sound). This introduces a new challenge to the language learning process. In French, for instance, it’s not particularly hard to spend a couple of weeks on pronunciation and then a couple of months on learning simple words using pictures. Simple words are, after all, relatively simple to learn, and once you’ve learned ~600 of them, you’ll find that it’s pretty damn easy to start devouring your grammar book, since you already know most of the nouns and verbs in it, and you just need to learn how to use them.

In Japanese and Chinese, however, learning simple words is actually fairly challenging. Here’s how you write ‘Artist’ in Japanese: 芸術家. All-in-all, you’re looking at 28 little lines, with a precise (and important!) order as to which lines you write first. Woof. Most textbooks will avoid this challenging bit by introducing you to the language through one of the phonetic alphabets that are easier to learn, teaching you some grammar and vocabulary in THAT, and then gradually introducing you to the logograms, in order of complexity. But this, too, has problems. It’s extraordinarily difficult to avoid constant translation, and you’re going to be buried in grammar at the same time as trying to wrestle with the writing system, which sounds like a real headache. It also means that you’re going to make spectacularly slow progress, since you have so many things to do at once.

Here’s where I’d like to land instead: I’d like to take care of the writing system ahead of time, so that by the time I open my grammar book, I already know ~600 words and their characters. That way, I’d be able to ignore most of the information about the writing system and instead learn to actually use my language. At least to me, this sounds like a much more comfortable learning process.

So what to do..clearly, starting directly with 芸術家 is going to be too difficult, but my alternatives (Heisig’s stuff, for instance) involve a great deal of translation, which I’m expressly trying to avoid. What I’m going to do instead is spend a week or two learning the most frequently used radicals – these are the building blocks of all of these characters – and then slowly working through my 625 word list, learning all component characters as I move along.

If you’re not familiar with the idea of radicals, here’s a quick introduction. When I look at the word “Apple”, I see a 5 letter word. What I DON’T see is ”A – 2 diagonal tall strokes, 1 horizontal mid stroke; p – 1 vertical low stroke, 1 curved stroke with an opening on the left, p – 1 vertical low stroke, 1 curved stroke with an opening on the left, l – 1 vertical mid stroke, e – 1 horizontal mid stroke, 1 most-of-a-circle-but-open-on-the-right stroke”. I see 5 letters because each of those symbols (A, p, l, and e) are chunked blocks in my mind. So I’m going to learn to do the same thing with Kanji/Hanzi characters.

Our 28-stroke word 芸術家 (artist), for instance, is made up of a bunch of little pieces. Take a look at the first character:


From the top down, it’s 艹, combined with 二 and 厶. If I knew those three characters in advance, then assembling this character (which means, roughly, ‘Art’) in my mind wouldn’t be that hard. It’s about as hard as learning a 3 letter word in any other language. My main job, then, is to make sure I know all of those component characters, so I can look at something as complex as 芸術家 and see a word composed of letters, rather than 28 individual pen strokes.

All in all, there are around 200 component radicals I need to care about. I can learn them all at once in the beginning, but it’d be more entertaining (and easier) to learn a subset of those components first (the very frequently used ones), and then start learning any new  radicals I need in the context of actual words. In practice, it’ll mean that I’m going to move through my 625 words relatively slowly compared with other languages, but once I get to the end, I’m going to be in a position to rocket through my grammar book, because the writing system isn’t going to hold me back. Also, it’s going to put me in a position to actually use resources like Google Images to improve my Japanese, where most people wouldn’t be in a position to do that until they’re most of the way through their grammar book. I’m going to start this process myself in October, and I’ll be posting about my progress here and on the blog. I’ll give you my preliminary plan here, but I’ll be tweaking it and improving it as I go, so expect some changes. If you start before me, and come up with some clever modifications or run into challenges and need some help, post them in the comments below so we can discuss them as a community and come up with some good solutions.

Without further ado, here’s my plan for Japanese: 

  1. Pronunciation: (Same as other languages)
    I plan to start with ear training and the Kana alphabets, and getting really comfy with those spelling systems over the course of 3-4 weeks. I’ll be using the Japanese pronunciation trainer, basically as soon as I get it ready for beta testers. If you’re not using my trainer, use these flashcard designs to get you going: Ear Training, Spelling.
  2. Radicals:
    Once that’s done, I’m planning on learning Kanji radicals. I’m planning on starting with the 48 radicals that make up 75% of the “Jōyō kanji” (the basic ~2000 Kanji required for literacy), which are listed in the beginning of this wiki article. I’m going to be using a lot of mnemonics in this language (see the Mnemonics chapters in the book!), basically choosing objects to connect to every single radical (some are already objects, like 口 (mouth) and 手 (hand), but some - 力 (‘power’, according to wiki), I might make a bit more concrete (‘energizer bunny’ or something). I’ll also put each of these radicals into Kanjidamage, and if he breaks them up even further, then I’ll learn those component parts as additional mnemonics. I’ve found from my Hungarian studies that you can have a practically infinite number of mnemonics and they never get harder to store, so there’s nothing lost by having lots and lots of them. I’ll be using Anki to memorize all of these, complete with their stroke order diagrams (basically instructions on how to write them out).
  3. 625 Simple Words:
    This phase resembles all the other languages, except I’m adding in a step where I learn all of the characters for these words, and whenever I encounter a character that has unfamiliar radicals, I’ll add new mnemonics for those radicals as well. For any words that have multiple characters, I’ll be using an extra couple of cards to quiz myself on the stroke order of each character individually, as well as that characters component radicals, and the sound (the on-yomi sound, for those keeping track) of that character when it’s used in that word. Kanji characters have an annoying habit of having all sorts of possible pronunciations (an issue that seems to be less problematic in Chinese), but there’s no reason why I can’t memorize the sound of each character  in the context of a given word. From a memory standpoint, the added bit of information (the sound) will make the character easier to remember, because I’m adding in one more association.
  4. Grammar and the language game: The rest of this process is the same as it would be for any other language. Once I get through these initial 625 words, I’m going to have a huge headstart on my Kanji, I’ll have my phonetic alphabets (Kana) down, and I’ll have a sizable vocabulary of really useful words. This will let me move through grammar at a relatively brisk pace. The rest is just play.

I’ll be posting an updated model deck in a few days. Basically, for 1-character words, I’ll have this:

  • Comprehension A: What’s 猫 mean? How do you pronounce it? [Answer: Picture of cat, Recording of "neko"]
  • Comprehension B: What’s ねこ mean? How do you pronounce it? [Answer: Picture of cat, Recording of "neko"]
  • Sound Production: Whats the word for [Picture of cat]? [Answer: Recording of "neko"]
  • Spelling Production A: What’s the Kanji for [Picture of cat] + ねこ? [Answer: 猫]
  • Spelling Production B (Optional): What’s the Kana for [Picture of cat] + Recording of “Neko”? [Answer: ねこ]
  • Structural Analysis (Optional, but generally you’ll want this on for at least the beginning): What’s the stroke order and component parts of 猫? [Answer: The full stroke order diagram + radical component parts]

For 2-or-more-character words, I’ll add more structural analysis cards, like this:

  • Comprehension A: What’s 列車 mean? How do you pronounce it? [Answer: Picture of a train, Recording of "ressha"]
  • Comprehension B: What’s れっしゃ mean? How do you pronounce it? [Answer: Picture of train, Recording of "ressha"]
  • Sound Production: Whats the word for [Picture of train]? [Answer: Recording of "ressha"]
  • Spelling Production A: What’s the Kanji for [Picture of train] + れっしゃ? [Answer: 列車]
  • Spelling Production B (Optional): What’s the Kana for [Picture of train] + Recording of “ressha”? [Answer: れっしゃ]
  • Structural Analysis (Optional, but generally you’ll want this on for at least the beginning): What’s the stroke order and component parts of _車 + [Picture of train]? [Answer: The full stroke order diagram of 列 + radical component parts + on-yomi sound in this context]
  • Structural Analysis (Optional, but generally you’ll want this on for at least the beginning): What’s the stroke order and component parts of 列_ + [Picture of train]? [Answer: The full stroke order diagram of 車 + radical component parts + on-yomi sound in this context]

Here’s a draft of my Chinese and Japanese model decks. There MAY be some bugs in them; if you see ‘em, report them in the forum and I’ll look into it. I’ll definitely catch any bugs once I start using them in late October!

Download the model decks for Japanese and Chinese