Hacker News Reader

"In Chinese, words are really compound words. For example, 熊猫 means panda. But 熊 means bear and 猫 means cat. This makes these unbounded morphemes, which is different than what we do in the Indo-European languages. And Chinese is composed of radicals that make things harder."

I'm intermediate in Japanese (conversational + able to read at an N2 level, albeit slowly), and my first response to anyone who says they "learned X kanji" is "how many words do you know?"

There's probably some value to learning the meaning of individual kanji in isolation, but it's very limited (especially in Japanese). Most words have only tenuous relation to the meanings of their individual kanji, and "learning kanji" is kind of a crutch for beginners to pretend that they're making rapid progress.

I suppose from that perspective, "learning X kanji in Y days" is a useful meme, because it pushes people to quickly move on from the approach. But that said, for most learners, it's much better to simply start learning words.

I disagree. It was impossible for me to just learn words, because kanji made no sense to me.

After RTK, I created ways to mentally breakdown and identify the characters. Then, suddenly, I was able to learn words much faster too - instead of struggling with “well this one has this curve here and then it points there and it means that”, reading became automatic, and I could focus on the word itself.

This has been my experience too. Like I said before, just learning characters doesn't let you read in of itself because there are compound words. But you have to think of characters as building blocks. Then you repeat the process with words and it is easier to form mnemonics because you already have the base structure. The idea is to start from the ground up and build a strong base before you move to building the house. This is opposed to the classical Western style that is building random walls first. The thing is that the latter will get you conversational faster, but the former leads to a better understanding but takes more time.

You have to learn kanji the same way you have to learn the alphabet. If you don't knoe the symbol you're going to have an extrenely hard time when you are learning writte. Word because suddenly you need to learn 3 kanji at once to understand a single word. Learning kanji and vocabulary at the same time isn't a big deal. At roughly 15 Kanji per day you can finish the first 2000 in 5 months but your vocabulary deck meanwhile is far away from being finished. I don't know why you are saying "move on" because as I said, it is one more thing you have in your toolbox. Your toolbox should consist of multiple anki decks(one of which could be a hiragana, katakana and a kanji deck, a listen only vocab deck, a reading only vocab deck for example), listening exercises, speaking exercises, grammar exercises and so on.

"You have to learn kanji the same way you have to learn the alphabet."

The RTK/RTH method is nothing like learning an alphabet: at no point in my education did we sit down for weeks and "learn the meaning" of A, B, C...etc. For phonetic alphabets like hiragana and katakana and roman and hangul, it's absolutely worthwhile to memorize the shapes of the characters, because there's a small set of them, and once you know them, you can convert written words to sounds. With RTK, that's not true at all.

An adult of average intelligence can learn to read kana in a week. That's time well-spent. Spending months memorizing the "meanings" of kanji is a whole other process, of dubious value.

"Learning kanji and vocabulary at the same time isn't a big deal."

This is true, and is the way that Japanese children learn the language: you learn a character, along with some words that contain the character. To the extent that individual kanji have meanings, you learn them in context.

"I don't know why you are saying "move on" because as I said, it is one more thing you have in your toolbox."

Most people who do RTK don't do it this way -- they do things like the OP, and try to "memorize X kanji in Y days". I think that's largely a waste of time.

> The three ticks on the left are the symbol for water but 十 means 10.

The right part in this case signifies a similar pronunciation:

汁 (zhi)

十 (shi)

This is a typical setup in Chinese character, part of it relates to the meaning, the other relates to the pronunciation.

The radical (which is usually on the left, or bottom) is sometimes semantically meaningful. In your example, 汁 has the water radical, and it is indeed a liquid. The rest of the character aside from the radical is usually entirely meaningless, aside from possibly providing the pronunciation.

As an example: 果裹猓粿蜾 are all pronounced identically - guǒ - but mean “fruit, package, monkey, rice cake, wasp” respectively. The radicals are more useful: in the last three, you have the “animal” radical (monkey), one of the “food” radicals (rice cake), and the “bug” radical (wasp).

How do you tell them apart in speech? Just context?

Usually they will form compounds with other characters, and the compounds will be unambiguous. For the first two characters, you have stuff like 水果、苹果（"fruit", "apple" respectively); 包裹 for "package". 裹 is also a verb by itself meaning "wrap":

> 我把身子用毯子裹上。

> Wǒ bǎ shēnzi yòng tǎnzi guǒshàng.

> I wrapped myself up in a blanket.

This last example shows that just like in English, words that are pronounced the same in Chinese can be distinguished by part-of-speech. You'd rarely confuse "red" the color and "read" (past tense), even though they're pronounced the same, because the first is an adjective and the second a verb (except in the old pun "What's black and white and read all over?", where you think "read" is actually "red" due to the grammatical structure of the sentence).

This is a pretty good explanation of how Chinese characters work, explained via English words: https://www.zompist.com/yingzi/yingzi.htm

(The other three "guǒ" characters are quite rare, so may be literary-only.)

> What's black and white and read all over?"

This is a riddle not a pun.

It is a riddle that uses a pun as a critical component to throw the person answering off the trail.

For sizzle.

I've explained in the article that I didn't concern myself with pronunciation at all which is the book authors advice. I have pronunciation on my answer side of the Pleco flash cards (they are shuffled and the algorithm decides which card to show me based on the ones I've marked as learned and their score) but I didn't make any effort to memorize it.

I focused entirely on how to memorize the characters and how to remember to write them based on the keyword + primitive.

> this book doesn't teach you how to read -- or speak -- but that doesn't mean it isn't useful

That's how chinese characters became a lingua franca between diplomats in ancient East Asia. They didn't have to speak each others languages, they just had to know the same character and communicate their ideas effectively enough on paper.

I've shared an excerpt from the book as an answer to @jmchuster below confirming this exact same thing.

RTK, which i'm assuming is similar to RTH, requires you to learn the cards in order because previous cards build upon the new ones. Do you mean you should shuffle your review cards? Is that not default Anki behavior?

I mean shuffle the review cards. Presenting them in order I think is part of the magic to these books (RTH and RTK share an author). It is the default behavior of Anki, if you add them all at the same time. It gets more difficult when you don't have all 1500 and start with, say, 100 and then build as you go (you don't want review time to be one month away for the first hundred characters just because you added them first). Because there is pronunciation and meaning clues to the radicals, the shuffle matters for long term memory. But the ordered helps get them into short term memory and to passively learn radical structures.

When I learned Japanese back in the day I did 35 Kanji a day with RTK + Anki and I made up all my own mnemonics.

It was an absolute grind. But it worked.