A few weeks ago I wrote about beginner Japanese. I’d managed to learn hiragana and katakana more easily than expected, with just a few weeks of effort, and in a moment of exuberant hubris, decided that my next project would be kanji.
Like most people, I always knew in the abstract that kanji was hard. There isn’t a fixed number of characters a proficient Japanese speaker needs to learn, but a basic set of 2,136 jōyō kanji is considered the bare minimum for functional literacy. At least a thousand characters on top of that are found in common everyday use, and many adults know a few thousand more. Coming from an alphabet of 26 characters, that’s a lot. Like, a lot a lot. Of course it’s hard.
But I only learned recently that kanji is actually way harder than that. Most kanji have at least two “readings”, which means that depending on the context they appear in, they’re read differently. Readings are not close together – in almost all cases, they’re completely different sounds that bear no relation to each other. On top of that, special rules like rendaku morph readings depending on context, adding even more cognitive overhead. As I’m studying, remembering what a kanji means is supremely easy compared to wracking my desperate, tortured brain for the alien sound associated with it when it’s surrounded by this or that other kanji/hiragana, often arbitrary-assigned to the point of cruelty. (It uses on’yomi reading usually for this, except that exceptional subset which uses kun’yomi instead, except for this exceptional-on-top-of-exceptional subsubset/word which goes back to on’yomi. And not that normal on’yomi reading, the other on’yomi reading.)
As an example, the kanji for “person” (人) is usually read as じん (ji-n) or にん (ni-n) when used in combination vocabulary, like 人口 (“population”, read じんこう or ji-n-ko-u) or 三人 (“three people”, read さんにん or sa-n-ni-n), but when it appears by itself as just “人” (again, meaning “person”), it’s read as ひと (hi-to). Other times it gets rendaku’ed, so “people” 人々 (“々” is a “repeater” that duplicates the kanji before it) is read ひとびと (hi-to-bi-to). And hey, for good measure it’s also sometimes just assigned a random alternative, like its use in “adult” (大人), read おとな (o-to-na) 1. Have fun with that gaijin.
And even if I’m able to build a decent kanji vocabulary, my next challenge would be actually reading it. So far I’ve only been examining kanji vocabulary in clean, convenient, completely unrealistic isolation. At some point, Japan decided that spaces were for losers, so when reading Japanese prose the mind operates like a regular expression engine, advancing as many characters as it can looking for a valid match, but not finding one, backtracking until matching a known word (even if the word is just one character). If you’ve been reading kanji since age five, your brain makes this happen automatically. If not, it makes you want to douse your Japanese textbook in gasoline, light it on fire, and launch whatever survives directly into the blazing hellfire of the sun. Yet another layer of difficulty in an already really, really difficult language.
The only saving grace is that the tooling available these days is amazing. I’m using WaniKani, which uses a spaced repetition system combined with custom-built memory mnemonics for every single radical, kanji, and vocabulary word, to help drill literacy into the brain, willing or not. Kanji advances levels from “apprentice” to “guru” to “master” to “enlightened” to “burned” as you’re able to successfully remember it after increasingly long periods of time. Forget it at any given level and it’s demoted, giving you another chance to work it back up. New kanji are unlocked over time as you as you make progress so that you’re not overwhelmed all at once. WaniKani’s web app works well, and it has a beautifully-built unofficial iPhone app (Tsurukame).
With tools like this, there’s never been a time in human history where learning Japanese has been easier 2, but that doesn’t mean it’s not still impossibly hard. I already feel right at the edge of my mental aptitude, and I’m only at level 3. WaniKani teaches you over 2,000 kanji and 6,000 vocabulary words over 60 levels, each of which can theoretically be completed in a week (in which I’m considerably behind schedule already). Even at breakneck pace, getting to level 60 takes more than a year, and probably more like ten for remedial learners like me. I fully expect to fall off the bus.
That said, it feels good for my brain, and it’s one of the more constructive things that I do in a typical day. Whenever a Japanese person tweets and I recognize a word, I get more excited than is even remotely reasonable. I’ll be keeping at it for now.
Did I make a mistake? Please consider sending a pull request.