Some character sets have very similar characters that are hard to tell apart while studying, near impossible at a later date.

For example, there's a lot of Japanese kanji, and they consist of Radicals (like common character fragments), meaning characters can contain the same radicals and literally be partly the same.

How can I differentiate these characters so they don't all get mentally jumbled together?

(While a majority of this would be done in handwriting and not-digital, respective digital typesets are applicable too)

  • Personally I prefer straight memorization based on lots of usage. Don't focus on similarities. Eventually mnemonic systems have an overhead that will get in your way.
    – user3169
    Commented May 13, 2016 at 0:24
  • 1
    @user3169 could you expand that into an answer? it has the makings of a great answer with a little bit more beef :)
    – Quill
    Commented May 13, 2016 at 0:39
  • Could you add some specific examples where you run into this problem? I generally don't care for mnemonic systems (for me it adds to the "mental jumble"), but its my opinion and results may vary depending on the individual and languages involved. As written your question may be too broad.
    – user3169
    Commented May 13, 2016 at 5:22

2 Answers 2


1. Don't learn similar characters together.

According to interference theory in Psychology, learning similar things together can cause those items to interfere with each other in your memory. So when you are shown, for example, "天", you may recall the meaning of the similar item "夫" if you learned these items at the same time. So avoid this as much as you can. If a book you're using puts two similar ones together, delay learning one until later.

2. Focus on distinguishing characteristics and use memory aids to distinguish them.

If you notice two characters are very similar, learn one first. when you learn the other, pay attention to what distinguishes them visually, and try to make an association with the distinguishing part. For example, 犬 means dog and 太 means great. So imagine the extra stroke in 犬 is the head of the dog, and you'll be able to remember the difference.

3. Write the characters as you learn them.

Don't just learn the characters visually; learn to write them too. Adding tactile information to your learning will aid your memory.

4. Learn radicals and pronunciation components well.

Try to learn the meanings of radicals and the pronunciations of characters that are used as the pronunciation component of a lot of characters. If you don't mix up the parts, you'll have less of a problem mixing up the whole. I don't know Japanese Kanji, but I'll give an example from Korean Hanja. If you know 古 is pronounced 'go', you won't mix up 若 (yak) and 苦 (go); it will be easily remembered that the second one is pronounced go, and the first one is not.

  • You might also add that sometimes you can make a determination based on expected usage in context. For example, once you learn 太陽 【たいよう】 in Japanese, it is unlikely you would write 犬陽 instead. The individual meaning of 太 vs. 犬 is not so important then, though "dog sun" would be meaningless. Also whether pronunciation follows similar character forms really depends on the language.
    – user3169
    Commented May 13, 2016 at 0:46
  • While controversial because it does not follow the standard, very rote, way of teaching the kanji, James W. Heisig's Remembering the Kanki touches upon the first 3 points here. He took great pains to learn some background on the kanjis, figure out radicals and find features that look similar and need to be distinguished, giving you his mnemonics to help you. More importantly, he teaches you how to do it yourself. This answer is the correct one in my opinion. If you need guidance on how to implement it for Japanese, look the book up. He does keep meaning and pronunciation for later studies. Commented May 28, 2020 at 13:11

I think people eventually get past this problem in Japanese by learning to read in context. Most of the time you're reading at the level of words, so you don't have to make that fine distinction to recognize something. It's more of a problem with handwriting. As long as you're inputting text by typing or speech recognition, the problem doesn't really come up.

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.