I finished RTK (Heisig, Remembering the Kanji) in May around six months ago, and today I decided to write this post as a reflection on what worked and what did not. I decided to do a kanji course because I found my ability to recognize and differentiate kanji was progressing much slower, prohibiting me from learning new words, and stopping me from progressing to the next level [1]. I couldn't decide which course to choose. Many on this forum choose Wakikani but I found it's pace a little slow, and I already spend eight hours a day on the computer for my job, and I wanted to spend less time on the internet, not more.[2]
I bought the book in early January and went slowly at first and made some mistakes. My first big mistake was deciding the mnemonics were silly tools for kids. This caused me to start over around 70 kanji in. Another mistake was setting my Anki deck to present me both keyword -> kanji and kanji -> keyword. Heisig states over and over why kanji -> keyword is the wrong approach, so I'll defer to his explanations. A third mistake was not having stroke order set up on my Anki cards. This Anki extension is helpful.
Eventually I settled on the following structure. The front of each card was the keyword and the back had the meaning and mnemonic on it. When the keyword came up, I would try to imagine the radicals in the kanji and the mnemonic, and from that I would physically write out the kanji in a notebook on my desk. I'd say this usually took around 40 minutes to an hour in the worst case. Then, I'd add 22 new cards to my deck (I also added radicals[3]), sometimes making my own mnemonics and sometimes borrowing or adapting them from here. While there are lots of pre-made decks online, I found creating my cards and creating my own mnemonics reinforced the meaning. Heisig encourages you to spend time thinking about good mnemonics, so adding 22 new cards sometimes took over an hour. Keep in mind that the better the mnemonic is, the less time you need to spend reviewing the kanji.
Even once I got settled into a routine around RTK there were many things I found difficult. After reaching around 800-900 kanji I began to start running into issues with synonyms. Keywords with the same meaning were difficult to get right. I never found a great solution to this, but I wasted a lot of time forcing myself to differentiating meaning between kanji with similar meanings. 簡 (simplicity) and 単 (simple) are a good example of a particularly difficult pair. Around kanji 1200 this had become seriously frustrating, so I decided that when I ran into a synonym collision, I'd put the other kanji on the front of the card. This made cards easier, reduced my review times a little, and helped me finish RTK. Forgetting mnemonics was also an issue. If found myself forgetting mnemonics during reviews, I'd often revise them to make them better. Mnemonics aren't set in stone. The other difficulty I had with RTK was the time commitment. RTK took me between an hour and an hour and a half each day. I have a full time job and social commitments so I found finding time for it in my life difficult. I resorted to waking up early regardless of when I went to bed, usually around 6am, and finishing RTK before heading to work around 8.
I finished RTK in late May around four months after starting it, and continued with Anki reviews for around a month and a half until I began frustrated that I was forgetting mnemonics. At the same time my reading ability was only getting better, and I started learning more and more vocab cards each day. One Saturday morning, a little hungover and frustrated about my inability to recall the mnemonics I had spent so much time making. In a fell swoop I buried all 2317 cards[2] in my RTK deck, and I'm glad I did that. What I got out of RTK was the system and something I can refer back to, but reviewing mnemonics until the end of time just doesn't seem that useful.
The main benefit of RTK is that I feel like I have a systematic way to understand kanji that lets me approach learning Japanese like a Chinese native would [4]. There is a part of my brain that exists now that did not before for kanji recognition and understanding. There's a good chance I could've developed this ability anyway if I just spent the two hours I spent on RTK every day doing something like reading, and looking up/writing down the kanji and words I don't know, then adding them to an Anki deck. This is how I learn new words and kanji today, and it works pretty well. However, back in January, I definitely did not have the confidence to do this, and RTK helped build that confidence and muscle to really engage in native sources that were previously too overwhelming.
One thing I thought RTK would help with but didn't is writing by hand. I almost never write kanji. My ability to write is very poor, and I don't think I can even write 勉強 from memory. For a little bit, I aimed to improve my writing ability by journaling daily and I found it difficult. I became frustrated that I could go from exertion to 勉 and strong to 強, but not be able to write 勉強 from memory, and I began doubting if RTK was a useful use of time. However, writing is not explicitly the goal of RTK, rather the goal of RTK is to develop a solid system for understanding kanji. "Character amnesia" is also not just a problem with Japanese learnings, but a general trend in the Japanese population due to the reliance on digital input systems. [4] We're not the only ワープロ馬鹿's out there. I don't think this is an issue, and I don't think it underscores the effectiveness of RTK much.
Overall, RTK took a lot of time and commitment, but I'm glad I finished it. It's of course hard to tell if there's something better I could've done with that time. I probably could've just spent the time I spent on RTK reading graded readers, learning words, or trying a more orthodox kanji course. However, I do genuinely feel like it helped me build a kind of structure around kanji that I did not have before, to the point where I can easily recognize and understand N2 and N1 level kanji that I come across, and I would recommend it to new learners.
[1] I see some people doing RTK when they don't know any Japanese. While the book can be used that way, I don't think I would've gotten through it when I was first starting out. I started RTK around N4 level.
[2] There were a few books I was deciding between, the main contenders were Kanji Look and Learn, Kanji in Context, and RTK. Kanji in Context looked promising and serious, but it states itself as being designed for intermediate learners and I did not think of myself as intermediate. Kanji Look and Learn seemed very accessible, but it just included N4 kanji and I preferred that RTK included all the kanji in order. I ended up deferring to RTK after reading a bit about Heisig, and while RTK is more unorthodox, I respected that it came from somebody with an academic background.
[3] Anki decks I see online for RTK don't do this, but I found it really helpful to also do spaced repetition on the radicals themselves. In order to get them into Anki (many of them aren't valid Unicode), I found a PDF of the text and screenshotted them, pasting the screenshot into the anki kanji field
[4] Character amnesia is an interesting phenomenon. I like this guy's short argument as to why it's probably not a big deal.
by Minolta-X700