How to Learn Kanji: All the Strategies You Need and More!

One definition of learning kanji is learning their meanings, readings, radicals, and stroke order. This is a tall order and it’s natural to wonder the best way to go about it. Learning kanji is a well documented challenge, particularly for non-native learners. It takes substantial commitment and dedication, and there is no one sure-fire way to do it, but with the right approach, it is entirely doable.

The best way to learn kanji is to employ a range of strategies such as component analysis, mnemonics, rote, and contextual learning. Effort must be made on both initial encoding into memory and retrieval practice. Systematizing one’s learning and consistent long-term practice will lead to success.

Kanji are logograms that represent both meaning and sound. There are different types which represent meaning in different ways. One common misconception is that all kanji are pictographs, that is, they are a picture of the meaning they represent. This is not at all the case. Most kanji are more complicated, at least on the surface. Furthermore, the sheer volume of kanji only adds to the learning challenge.

Fortunately, there are lots of learning strategies at your disposal. With a bit of background knowledge about kanji and an awareness of learning strategies, you will be well equipped for the kanji learning mission. Let’s first discuss the background, then we’ll get into the strategies, and we’ll wrap up with how to systematize your learning.

There are a few different ways to classify kanji into types. For practical purposes there are four types of kanji, which are the first four listed below. But a common classification is the into six types known as the 六書 (rikusho) [1][2], these are the first six types listed below. But a seventh type of kanji also exists called 国字, these are characters that were developed and used solely in Japan.

Types of kanji:

象形文字 (shoukeimoji) Pictographic – Pictorial representations of meaning
指示文字 (shijimoji) Ideographic – Representations of abstract concepts such as numbers and directions
会意文字 (kaiimoji) Semantic composite – Combinations of two or more kanji forming a single kanji
形成文字 (keiseimoji) Semantic-phonetic composite – Combinations of two parts of which one is suggestive of meaning and the other represents sound.
転注文字 (tenchuumoji) Derivative kanji – Kanji that have been derived from older kanji, but for which the modern meaning is now dissociated.
仮借文字 (kashamoji) Loan kanji – Kanji that are use for their phonetic reading only and not their meaning.
国字 (kokuji) National kanji – Kanji that were made in and used solely in Japan

Note: It is really the first 4 types that you want to be the most aware of. An argument can be made that all derivative kanji, loan kanji, and national kanji can fall into at least one of the first 4 categories.

The various types of kanji are not equally common. The list below provides a detailed breakdown of percentages of the 2136 jouyou kanji [3]. It turns out that sematic-composite and semantic-phonetic composite types make up the vast majority. This is important because a lot of your kanji learning strategy will be geared towards learning these types of kanji.

Percentage and amount of each type of jouyou kanji:

Pictographic – 12.4% (265 kanji)
Ideographic – 0.5% (10 kanji)
Semantic-composite – 24.6% (530 kanji)
Semantic-phonetic composite – 61.4% (1312 kanji)
Derivative kanji – n/a
Loan kanji – 0.6% (13 kanji)
National kanji – 0.3% (6 kanji)

Yet another important note is that kanji do not all appear with equal frequency. Some are extremely common, while others are more rarely encountered. In Japanese printed materials, about 200 high-frequency characters account for 50% of kanji usage, 500 account for 75-80% kanji usage, and 2000 account for 98-99% [4][5]. This is why you may have heard that you need to know about 2000 kanji to be literate in Japanese.

Kanji Learning Strategies

In order to learn kanji as effectively and efficiently as possible, it is important to realize that there is no one strategy that can be applied universally. A number of different strategies are needed for success. In fact, research indicates that the most effective kanji learners use a wide variety of strategies [6][1]. Let’s look at what strategies are available to us.

Strategies for learning kanji:

Pictorial association
Component analysis
Mnemonics
Phonetic association
Rote learning
Contextual

All these strategies have their place in your kanji learning arsenal, but each needs to be used for a specific purpose. This leads us to the important point that learning kanji involves two processes: encoding and retrieval practice. Encoding involves integrating new information into memory. Retrieval practice involves frequent retrieval of this new information to strengthen the memory trace [1].

The first thing to focus on is encoding. The better you encode kanji into your memory, the easier it will be to remember. We will review each encoding strategy below. They are all different, but the common thread between them is connecting new information with known information, either your worldly knowledge or the kanji knowledge that you will accumulate once you start learning.

These strategies will help create powerful links in your memory, but another important key to keep in mind is that different strategies are optimal for different kanji types [1]. Professor Heath Rose, who wrote an entire book on learning kanji, says the following:

…by using strategies that are best suited for each kanji type, the learner can greatly reduce the cognitive energy spent on kanji learning.
Heath Rose

Pictorial Association Strategy for Learning Kanji

Pictorial association is a highly effective strategy for learning shoukeimoji (pictographic) kanji. Since the kanji are essentially pictures of what they represent it is easy to make the visual link. Almost half the kanji taught in first year elementary school in Japan are pictographic [1]. The case is similar for many beginner level textbooks for foreign learners.

Furthermore, many teachers and textbooks will show how kanji such as 山 and 川 actually look like what they represent, mountain and river respectively. So pictorial association is a common strategy for beginner kanji learners and is quite helpful early on. But since only about 12% of kanji are pictographs, this strategy’s usefulness is limited.

It’s not a good idea to use pictorial association for kanji that are not of the pictographic type. The reason is that you will probably look at the kanji, think that it vaguely looks like something, but if that something has nothing to do with the meaning of the kanji, it will make for a poor encoding that is difficult to remember.

Component Analysis Strategy for Learning Kanji

…component analysis is a strategy that many learners embrace consciously, and that research shows to be absolutely essential to progress to the advanced stages of Japanese written proficiency.
Heath Rose

Component analysis is arguably the most important strategy for kanji learning. Most kanji are not unique logograms, but rather a composition of two or more sub-components. These sub-components are the building blocks of kanji. Component analysis is all about knowing these building blocks and analyzing how they come together to form a kanji.

There are close to 500 components that make up the 2136 jouyou kanji. For example, Kanshudo provides a great list of 439 components that each appear in at least 4 jouyou kanji. As this suggests, like whole kanji, components do not all appear with the same frequency. For example, the water component (radical), 氵, is high frequency and used in 110 of the jouyou kanji.

It’s important to pause here and mention that the terms ‘component’ and ‘radical’ are often used interchangeably, which will be the case in this article. But, in the strictest of definitions, not all components are radicals. Technically there are only 214 historical radicals (and not all of them are used in the jouyou kanji) [7].

Historical radicals are root components used to organize and group kanji with others containing the same radical. All kanji contain a historical radical with semantic properties that hint at the meaning of the kanji [7]. For example, 金 is both a radical and a whole kanji on its own. Alone it means gold. But when paired with other radicals it can form kanji such as: 銀 (silver), 銅 (copper), and 鉄 (iron).

For the most part, radicals (components) are placed in predictable positions relative to other radicals within a kanji. Based on their positioning, radicals fall into seven broad categories (although this can be expanded to up to 12 categories).

Categories of radicals:

Hen 偏（へん）– positioned on left side of a kanji
Tsukuri 旁（つくり）– positioned on the right side of a kanji
Kanmuri 冠（かんむり）– positioned on the top part of a kanji
Ashi 脚（あし）– positioned on the bottom part of a kanji
Tare 垂（たれ）– wraps around the top part and left side of a kanji
Nyo 繞（にょう）– wraps around the bottom part and left side of a kanji
Kamae 構（かまえ）– wraps around the outside of a kanji

Component analysis strategies are probably the most widely applied strategies at the intermediate and advanced stages of Japanese language learning [1]. Being able to decompose kanji into simpler components makes kanji less intimidating and is a great aid for memorization. But the real power of component analysis is when it is used in concert with other strategies, such as mnemonics.

Mnemonics Strategy for Learning Kanji

There is strong evidence that a mnemonic strategy provides a powerful tool to link kanji and kanji components to their meaning.
Heath Rose

A mnemonic strategy can be defined as the use of stories or phrases to link kanji to the meaning and sounds they represent [1]. And as mentioned, mnemonics often go hand in hand with component analysis. By knowing and naming radicals, you can make a story about the radicals that make up a kanji which link to its meaning and reading.

As a simple example, lets take the kanji 字 (じ), which means ‘character’ or ‘letter’. Let the top radical, 宀, be called ‘roof’ and let the bottom radical, 子, be called ‘child’. Note, in general radicals have their own meaning or name, which are what we are using here, but you are free to make up your own as well. Now we need a story to connect how a roof and a child mean ‘letter’ and be read as じ (ji).

In general, children learn to write in school, and a school is a building with a roof. So, our story can be “under the school roof, children learn to write letters, and the letter they struggle to write the most, is the capital G (ji).”

The danger of using mnemonics is when it becomes excessively difficult to relate components together in a logical way, or the story becomes so complicated that it is hard to remember. Either way, you may get lost in the mnemonic and lose the end goal of trying to arrive at the kanji’s meaning and reading.

Phonetic Association Strategy for Learning Kanji

As the name of the strategy suggests, phonetic association is all about tying a kanji to its reading. The natural target of this strategy are keiseimoji (semantic-phonetic composites). With this kanji type, the phonetic component may be useful in deriving or retrieving the onyomi reading of the whole kanji in which it appears [8]. Note that the phonetic component may comprise of more than one radical.

As mentioned above in our component analysis discussion, components are often positioned predictably within a kanji. By knowing the relative positioning, the phonetic radical can often be identified. The diagram below shows the relative positioning of phonetic and semantic radicals.

Positions of semantic and phonetic radicals within kanji (Chikamatsu 2005, [5])

It is important to mention that phonetic association needs to be used judiciously and a few disclaimers need to be made. One is that the real power of this strategy is when trying to recall already learned kanji. The phonetic component and it’s sound can serve as a memory anchor point or trigger.

Another disclaimer is that kanji often have multiple readings, so when trying to read a word written in kanji, the phonetic component won’t necessarily tell you which is the correct reading of the kanji for that specific word. This is a key point. The strategy is more useful at the individual kanji level for learning one onyomi reading for a given kanji, but at the word level, the strategy can breakdown.

In a research paper entitled “Identifying Useful Phonetic Components of Kanji for Learners of Japanese”, researchers sought to answer the question: for which of the phonetic components is it useful to learn one onyomi reading (pronunciation) and be able to rely on that one pronunciation for reading and recalling kanji that use those components. They applied a set of four criteria to identify the most useful components.

Learning a phonetic kanji component is useful if:

It occurs in commonly used words,
has only a few pronunciations (onyomi readings),
the reliability of one of the pronunciations is high (i.e., far more common than the others),
and there are multiple kanji representing the same pronunciation.

In the end they found 69 highly useful components and an additional 50 relatively useful components for a total of 119. Both groups of components are given below.

Highly useful phonetic components of kanji (Toyoda et al. 2013 [8])

Relatively useful phonetic components of kanji (Toyoda et al. 2013 [8])

Important Points about Kanji Readings

Let’s pause here and talk about kanji readings (which above have also been referred to interchangeably as sounds or pronunciations). As mentioned, in most cases kanji have multiple readings. Of the readings, there are two types: onyomi (Chinese origin reading) and kunyomi (Japanese origin reading). When first learning and encoding kanji, I would recommend you learn just one onyomi reading and one kunyomi reading.

Onyomi readings are often more numerous than kunyomi readings and are mainly used for jukugo (compound-kanji words). Kunyomi readings are mainly used for verbs and words represented by one single kanji on its own. For onyomi, you will learn a reading that will later be combined with the readings of other kanji to form compound words.

For the kunyomi reading, I recommend always learning the kanji with a whole word, usually a verb or the single-kanji word if applicable. For example, lets take the kanji 霧, which means fog. It has one kunyomi and three onyomi readings. The kunyomi, which is the reading it takes when on its own, is きり (kiri). For the onyomi reading, we can take む (mu) as the one reading we’ll memorize.

The む reading is used for example in the yojijukugo (four kanji compound word) 雲散霧消 (うんさんむしょう) which means ‘vanishing like mist’. 霧 is a good example of the complexities of kanji readings because when combined with other kanji to form compound words, depending on the word, it might use one of its onyomi readings or its kunyomi reading. For example, the word 霧雨 is read きりさめ and means ‘light rain’.

Summary of Kanji Encoding Strategies

Let’s summarize the discussion of using the right encoding strategy for the right kanji type. Leverage a pictographic strategy for shoukeimoji (pictographic kanji), but don’t use it for other types where what you think the kanji looks like is not in any way related to the kanji’s meaning. Use component analysis for any kanji containing multiple components, namely kaiimoji (semantic-composites) and keiseimoji (semantic-phonetic composites).

Use mnemonics with component analysis as much as possible to tie components together to arrive at the meaning and reading of a kanji. And use mnemonics for shijimoji (ideographs). Leverage phonetic association where useful to anchor your memory around the phonetic component of keiseimoji.

Kanji Memory Retrieval Practice

Memory retrieval practice is something you need to incorporate into your learning regimen almost as soon as you start learning kanji. Particularly as you learn higher and higher numbers of kanji, this becomes critically important. One of the big challenges with kanji is simply remembering the ones you’ve already learned. Retrieval practice involves both recognition and production practice.

Just as the word suggests, recognition is about recognizing the meaning of a kanji when you see it and being able to read it. But there are two layers to recognition. The first is just recognizing the individual kanji, recalling its meaning, and the onyomi and kunyomi readings that you memorized initially. The second layer is the more practical layer, which is being able to read kanji as part of words in a text.

Flashcards are a great way to practice both layers of recognition. You can make cards for just kanji, and you can make cards with vocabulary written in kanji. In fact, when learning kanji, you should also be simultaneously learning to read vocabulary that uses the kanji you are currently learning. This could be words you already know, but for which you didn’t know the kanji, and new words as well.

One of the key benefits of learning kanji as part of words is learning the other readings associated with a particular kanji that you did not learn initially. In fact, as we touched on above, there is really no reliable way to predict the reading of a word written in kanji if the kanji in the word have more than one reading, particularly more than one onyomi reading.

Although this might seem like an impediment, but the reason it is actually a benefit is because you don’t need to memorize a bunch of additional readings for individual kanji in advance, rather, when you learn the word, you learn its reading as a whole and get the individual kanji readings as part of the deal. Let’s further extend this idea of learning kanji in the context of words.

Contextual Strategy for Learning Kanji

One of your goals for learning kanji is probably to be able to read in Japanese. Recognition of individual kanji and words written in kanji is one of the pre-requisites for reading. Learning kanji and learning to read are highly interrelated and complementary. That is, learning kanji will help you to read and reading will help you to learn kanji.

Practicing reading and learning kanji through the context of their use in words is called a contextual strategy. With enough practice and exposure over time, you begin to recognize whole words written as kanji compounds or kanji-hiragana along with their meaning, reading and the broader context of the situations in which they are used.

A contextual strategy (or simply, just reading) has the multiple benefits of training your kanji recognition ability, your kanji word reading ability, and building your vocabulary. And a great companion to a contextual strategy is again, flashcards. When you come across words you don’t know, you can look them up and store them as flashcards for later review, for example using Anki.

Rote Learning Strategy for Learning Kanji

Now on to production. This is where rote learning, writing out kanji over and over, comes into play. There is a big difference between being able to recognize kanji and being able to produce (write) it. Unsurprisingly, to learn to write kanji, you need to practice writing it. Professor Rose contrasts the functions of mnemonics (encoding) and rote learning (retrieval) in the following quote:

From a cognitive theory point of view, we understand that mnemonic strategies are used to enhance encoding of new information and rote learning retrieves this encoded information.
Heath Rose

Unfortunately, rote learning is often used as the first and only strategy for learning kanji. In fact, it will probably not surprise you that several researchers have reported rote learning to be the most common strategy used by learners and the most recommended by Japanese teachers [9]. But in terms of initial encoding, rote learning is a far from an optimal strategy.

If you were to just use rote learning, it’s hard because you will be learning kanji as an ordered pattern of a specific number of strokes, rather than an assemblage of components tied together with a mnemonic, for example. Ignoring smaller and more easily remembered components and memorizing kanji by just trying to draw out the correct pattern of ordered strokes is very error prone and hard to remember.

But by learning components and doing an initial encoding of kanji first, it will be much easier to reproduce with the correct stroke order and count. With enough practice, writing kanji will become part of your procedural memory and you’ll be able to write it more naturally [1].

Systematizing Your Kanji Learning

…the coordination and management of the study of written Japanese is probably even more important than the initial encoding itself for long-term success in reaching a functionally literate level in Japanese writing.
Heath Rose

Learning 2000+ kanji is a big undertaking. The best thing to do is to commit to the long term and set aside time on regular (preferably daily) basis to study. Indeed, make kanji learning part of your daily routine. With your time set aside, as the quote from Professor Rose above alludes to, you need a system to manage your learning. Those that plan and systematize their learning are the most successful [1].

At the most basic level, your kanji learning system needs to include incrementally learning manageable numbers of new kanji, reviewing previously learned kanji, and tracking your progress. You could make your own homemade system. For example, to manage and track your learning, you could make a big spreadsheet of the 2136 jouyou kanji and learn them in the same order they are taught in Japanese schools.

Here are lists of the first 1006 kanji, the subsequent 1130 kanji, and components from Kanshudo. You will need to work on encoding the kanji you learn in the right way, and you will also need to set up a review schedule for practicing recognition and production of already learned kanji. In terms of your review schedule, you are going to want to leverage a spaced repetition system (SRS).

Spaced repetition is the way to settle information in long-term memory [6]. It essentially involves review at time intervals that get progressively farther apart. When you forget a kanji or make a mistake with one, you mark it for more frequent review. Again, flashcards will be helpful. Anki is a popular option that employs SRS.

Completely making you’re your own system is a task in and of itself. Fortunately, there are many premade kanji learning apps and online services available that you can take advantage of. These apps organize your kanji learning by presenting new kanji to you in a set order, and often group kanji in levels. They test you regularly on learned kanji, often via SRS, and track your learning progress.

Some even have a feature to have users draw kanji on a device’s screen by finger or stylus and can determine if the correct stroke order and shape were input. One app like this that I have enjoyed using is Kanji Box. For your kanji learning system, I would highly recommend you use one or more existing apps and services.

My Kanji Learning System (that you can use too!)

As a starting point, let me tell you my learning system. I use Wanikani as my primary guided learning and tracking tool. It’s a great tool that incorporates mnemonics, component analysis and pictorial association (where applicable) into its teaching system. It presents kanji and components in a logical order, employs SRS and tests you regularly.

Since Wanikani only tests recognition and does not include a feature to test production of kanji through onscreen input, I’ve supplemented this by inputting all the kanji and radicals in Wanikani into Anki flashcards. One side of the cards has the kanji meaning, the other has the kanji and one onyomi reading. I set my Anki SRS to be the same as Wanikani’s and set the cards so that I will be shown the meaning side first.

When shown the meaning, I then try to correctly write the kanji and remember the onyomi before turning the card. I also try to write with the correct stroke order. In fact, since I have already employed more effective means to encode the kanji into my memory, knowing the stroke order is actually very helpful to pull it out of my memory (even though I said it was error prone above). Here is one place where you can get kanji stroke order diagrams when learning for the first time.

So now we have covered the kanji mission of learning the meanings, readings, radicals and stroke order [6]. But now you will be asking, how long will it take and how hard is it going to be? Why is kanji even important? Why does is continue to exist in Japan? I have written about all these topics. Please see the links below.

Even more articles about kanji!

References:

[1] Rose, Heath. (2017). The Japanese Writing System: Challenges, Strategies and Self-regulation for Learning Kanji. Multilingual Matters, Bristol, Blue Ridge Summit.

[2] Paxton, Simon. (2019). Kanji Matters in a Multilingual Japan. The journal of Rikkyo University Language Center, 42.

[3] 山本, 康喬. 漢字の分類（象形、指事、会意、形声、仮借）と音符. 第３回音符研究会. ２０１７年５月１３日. 緑のgoo.

[4] Mori, Yoshiko. (2012). Five Myths About Kanji and Kanji Learning. Japanese Language and Literature, 46, 1.

[5] Chikamatsu, Nobuko. (2005). L2 Japanese Kanji Memory and Retrieval: An Experiment on the Tip-of-the-pen (TOP) Phenomenon. Second Language Writing Systems, Chapter 2. Multilingual Matters.

[6] Brasil de Sá, Michele Eduarda. (2015). Studying How To Study Kanji: A Practical Approach. University of Brasília. Asian Conference on Language Learning.

[7] Lory, Harumi Hibino & Bosse, Arno. The 214 traditional kanji radicals and their variants. Kanji Alive. The University of Chicago. (Accessed 2023).

[8] Toyoda, Etsuko et al. (2013). Identifying Useful Phonetic Components of kanji for Learners of Japanese. Japanese Language and Literature 47(2).

[9] Gamage, G. Haththotuwa. (2003). Issues in Strategy Classifications in Language Learning: A Framework for Kanji Learning Strategy Research. University of Wollongong.