Language Cities, Gaining More, and Applications for Writing and Pronunciation

Hey everyone!

I’ve been slowly working my way through Japanese in my spare time for a while now, and now that I’m getting deeper into memory training, I’m looking to using both disciplines to help the other. I’ve been reading through Dominic O’Brien’s How To Develop A Perfect Memory, which details his method. Although I imagine a great many of you are already familiar with it, for any who aren’t, I’ll outline it briefly. If you already know it, feel free to skip to the next paragraph. It boils down to not organizing words in a journey format, but within a city or town that you know well. When learning a word, for example “(to)turn.” You find a location that you (hopefully immediately) associate with the word “turn” in that town. In my case, this is a particular intersection where I always make an unprotected left turn. When you think of the word, you think of that place, which gives you a place to put an image. The image then is made from what the foreign word sounds like. In Japanese, this is “magaru.” So when I say that out loud, it makes me think of margaine and Roo from Winnie the Pooh. So I picture a huge melting stick of margarine in the middle of the intersection. On top of it is the Land O’ Lakes lady (because I associate her with margarine) and she’s fighting with Roo. And so for each language, you learn by doing this with each word, sometimes separating different kinds of words in different areas of town based on grammar.

From this, some…not so much issues, as points of query have arisen. With Japanese and Chinese in particular, the writing system is rather complex, and adds another level of memorization on top of the sound of the word. My method in the past, which has worked, even when not applied to loci, is to create an image based on the first two radicals of a kanji/hanxi character. However, the unaddressed point by O’Brien (at least in this book) is about spelling memorization, particularly with writing systems so vastly different from English. I’ve been working to put an image based on the characters of a word next to my image for the pronunciation, and having them react as much as I can without muddying things up too much. I’ve sort of gone back and forth so far on whether to use my radical-based images or images based on the meanings of the characters themselves. For example, the character 揮 means (among other things), to wave. But my radical mnemonic for it is a hand in an army glove, because the left radical means hand and the right means army. Maybe as long as I strongly associate that image with the character and don’t lose the meaning “to wave,” in there it won’t be a problem, but it does make me wonder.

There is a lot of information packed into each character, much less words with multiple characters. And so trying to place both a pronunciation image, and an image for the written word based either on individual character meaning or individual character radical-produced imagery is a lot to cram into a location in the first place, not even factoring in a hopefully fairly speedy recall. And then there’s the matter of storing the various meanings that a particular character can have. Often times a single character will have an inherent association or meaning; for example 食 has the association of food, but the actual word for food is 食べ物. Do I reserve my food location for the actual word for food or the character that has that association (my inclination is for the former, but this is for argument’s sake)? And then where do I put the information for the one not placed at that location? It seems like things could spin out of control very quickly and end up with me forgetting at least one of those many pieces of information.

I’m curious how others approach language learning, especially with a language like Japanese.

On a different, less complex note, I’d like to learn more languages as well (hopefully ones not quite as complicated to start with as Japanese, hahah), but have only lived in so many cities and have so many areas with which I have many associations to serve as loci. Does anyone have any tips for an intuitive way to familiarize oneself with a city or city surrogate without necessarily living there and going through everyday life? My instinct is open-world sandbox games with detailed varied environments. This could be both fun and helpful, mnemonic-wise. Unfortunately my experience with these kinds of games is mostly limited to Batman: Arkham City, which doesn’t exactly hold the MOST varied locations in the world. Does anyone have any recommendations for games that could serve a purpose like this or other methods in a similar vein?

If you read all of this, thank you very much, and I’m sorry for the essay-length entry!


japanese doesnt have many syllables so it should be possible to make an image for each kana letter. It takes more images to encode words but its much less ambiguous. Have you tried doing something like that? If you have a number system you could assign each number to a syllable like 01,02,03,04,05 = a,i,u,e,o and then 06, etc = ka, etc. I don’t know if that works in practice but its an idea I had.

That’s definitely something I’ve thought about doing, since sometimes coming up with images by soundalike is easy, but other time I feel like I really have to reach for it, or just come up start coming up with images for syllables, like Wario for “wa.” I may get deeper into doing that and use it when a natural image won’t make itself apparent.

I’ve been thinking more about where to store individual characters in the grand scheme of things, compared to how I store and recall words. The guiding light so far has been how I’ll need to recall the information. So far, since I’m encoding words by locations related to the word’s meaning and storing images for pronunciation and kanji usage there, I’m most easily setting myself up to remember the words when I’m prompted to recall what an English word is in Japanese. My hope is that this also translates to when I hear a word pronounced and am having trouble recognizing it, it will call to mind the soundalike image which will get me to the location. Time will show the efficacy or weakness of this.

But where individual kanji are concerned, it’s less likely that I’ll be asked to recall a random single kanji out of the blue, so this same method may not be necessary for individual kanji that don’t see a lot of solo use. So when will I need to remember an individual kanji’s meaning the most? To my mind, it seems mostly likely to come up more often when seeing it, whether online, in a newspaper or book, or in real life. So I need the character itself to conjure a memorable image that relates to its meaning. As I discussed in my original post, I’ve systematically gone about (and still am going about) learning the radicals used in kanji and using them to create a memorable image or image story. So take, for example, the character 爆. On the left is the radical for "fire. On the top right is the radical for “sun,” below that, the radical for, “together,” and below that the radical for, “water.” So my mental image for this character is Calcifer, the flame demon from Howl’s Moving Castle joining hands with an anthropomorphic Sun (together) and jumping into a pool of water. So seeing the character will hopefully make me think of that story. From there, all I need to do is link the meaning to that image. As with most Japanese characters, this one has multiple, though related meanings: bomb, burst open, pop, split. So maybe Calcifer and the Sun jump into the pool, which causes a big explosion to rise from it. Maybe it’s an underground pool so it bursts open. The two fall out of the pool, and Calcifer pops like a balloon, and the Sun splits in half down the middle. Or something like that.

So my theory, and what I think I’ll try, is to make relate words to locations that will contain images to sugges pronunciation and the kanji used, and relate individual kanji and their meanings through visual recognition of their radicals and linking the meanings to the images those produce. I have no idea if this is a good or effective method, but where my mind is right now, it seems to make sense. I’d love to hear your thoughts, Grandzam, and anyone else who wants to chime in! Thanks for taking the time to read and respond :slight_smile:

Hey Chris,

I fully understand your problem. Ive lived in Japan for 25 years and learning vocabulary especially kanji has been a uge stuggle and I have long puzzled over the correct approach to this. I dont think the loci method works here due to the huge number of even basic kanji involved (the 2000 plus joyo)

So I suggest this.

  1. Download the JOYO kanji in ANKI (just learn the meanings in English)
  2. Make a story based on the radicals for each one you dont know which you can recall easily.
  3. Note this down in a file and keep it.
  4. Rinse and repeat.
  5. Create another ANKI file of your own, with the kanji on one side and the meaning and story on the other.
  6. When you have finished the first ANKI deck, continue to revise using your own ANKI deck (See 4)

If you do this and use ANKI to revise constantly you will master the kanji pretty quickly (3-6 months). From here you can go on to learn the readings and kanji in combination using the same techniques.

Warning, don’t stop using the ANKI file just because you have finished. You need to use it every day for a year or more to get this shit thoroughly into your head. Look at the kanji, try to remember the story associated with it and allow this to lead you to the meaning. You need to remember the story, not just try and jump straight to the meaning. Resist this temptation. If you don’t remember the story you will forget how to get to the meaning. Also you must read in Japanese as well (I suggest bi-lingual books) to reinforce this.

When dealing with huge volumes of chinese characters, and a hugely different vocabulary, you need to use a combination of methods, basically this is the Heisig method hooked up to an ANKI engine. There is no perfect way to do this, believe me I have been pondering this for over two decades. But this method does work and its actually really quick and thorough compared to many others. Once you have the JOYO under your belt, the whole language starts to open up to you. The major problem with Japanese and Chinese is not the grammar or pronunciation etc., its the fact you cannot read that bogs down the whole enterprise of language learning because this hinders the aquisition of vocabulary.

Good luck and ask me anything you need to know.