I got into learning Chinese recently and since I already speak Japanese I find myself in a weird spot of being fairly familiar with a writing system, yet barely being able to read more than 30 characters. As for Japanese character readings, I learnt them using radicals that indicate or hint a certain reading. According to my Chinese friends, they too read Hanzi by radicals. But since the readings differ largely, Japanese has more possible readings per character than Chinese and some characters are flat out different, I obviously can't apply the same system again. Does anyone know of a book or a website that teaches the readings by radicals? (Something like Heisig's Remembering the kanji 2)

  • Consider using 《汉字字根》, this book is written by 张建铭 & 张婉如(JianMing Zhang & WanRu Zhang).
    – xqMogvKW
    Nov 27 '16 at 4:47
  • Japanese has a concept of radicals because the verbs, adjectives will change its look in accord with a certain rule and learning radicals will help construct your sentence in different occasions(e.g. honorific, past tense, imperatives). However, Chinese is not a language of radicals since its vocabularies will never change its form in any occasion! We use different vocabularies, not different conjugation rule, to identify the occasion. Nov 27 '16 at 8:00
  • @超酷超帅型头型男 I think you got me wrong there, I wasn't referring to particles or conjugations but radicals the Kanji and Hanzi respectively are composed of. I am well aware of the fact that grammar is very different however that wasn't my point, I was solely looking for a method to read Hanzi by radicals (here's an example of what I meant, in Japanese every kanji containing the radical 中 shares a common reading. Kanji that contain this radical include 仲、沖、忠 and they all share a common reading). Thusly I was looking for a book/website listing these.
    – BluNova897
    Nov 27 '16 at 13:35
  • @DongWei thanks for the suggestion, since I am planning on studying Taiwanese mandarin, I was wondering if there is an equivalent for traditional characters?
    – BluNova897
    Nov 27 '16 at 13:38

It is possible in Chinese as well, and similar rules apply.

The big reason behind this is the liushu 六書, or the six methods by which all Chinese characters are created: pictogram (象形), simple ideogram (指事), compound ideogram (會意), rebus (假借), phonetic-semantic (形聲), and transformed cognate (轉注). The method of reading by radical is only applicable to the phonetical-semantical category. However:

  • pictograms and simple ideograms make up a small (although very frequently used) portion of Chinese characters due to the poverty of these two methods in conveying relatively complex ideas. Pictograms and simple ideograms are one-piece in their own, and have no phonetic and/or semantic radicals;
  • rebus and transformed cognate characters are rare, and their origins relatively obscure;
  • a good percentage of Chinese characters are compound ideograms (in Japanese kanji, virtually all kokuji are compound ideograms). However, radicals for compound ideograms usually carry no phonetic value.

Fortunately, the rest, or phonetic-semantic characters (形聲字), make up the vast majority (>70-80%) of Chinese characters. So, yes, you may read them by radicals (but very carefully). However, in Chinese, the onset (聲母) and tone of many 形聲字 have changed, so their onset and tone is often different from the onset and tone of their phonetic radical (聲旁). Let's take characters in the phonetic series 方 as an example:

  • 方 "square": Standard Mandarin fāng, Standard Cantonese fong1, Japanese go'on hō (ホー), Japanese kan'on hō (ホー). Reconstructed Middle Chinese (according to Pan Wuyun 潘悟云) /*pʷiɐŋ/, reconstructed Old Chinese (according to Zhengzhang Shangfang 鄭張尚芳) /*paŋ/.
  • 訪 "visit": Standard Mandarin fǎng, Standard Cantonese fong2, Japanese kan'on hō (ホー). Middle Chinese (Pan) /*pʰʷiɐŋH/, Old Chinese (Zhengzhang) /*pʰaŋs/.
  • 旁 "side": Standard Mandarin páng, Standard Cantonese pong4, Japanese go'on bō (ボー), Japanese kan'on hō (ホー). Middle Chinese (Pan) /*bɑŋ/, Old Chinese (Zhengzhang) /*baːŋ/.

We see that reading by radical usually works in Japanese, but can be tricky to apply in Chinese. As a rule of thumb, use radicals to help remember the rime (韻母) part only. You still have to memorize the onset and tone. Moreover, since many of the onsets have already changed in the Old Chinese era, so really memorized the onsets when reading by radical.

Even with Japanese, reading by radical is not entirely bullet-proof. It can't help you distinguish go'on vs kan'on (which can be tricky), and it can't help you decide when to use on'yomi and when to use kun'yomi (also tricky).

PS: Standard Mandarin refers to Putonghua 普通話 and Kuo'yu 國語, which are both standard registers of Mandarin based on the dialect in and around Beijing. 普通話 is the standard in mainland China and 國語 is the standard in Taiwan: they are essentially the same language, with some disagreements on the pronunciation of a few words. Standard Cantonese is the de facto lingua franca in the Cantonese-speaking world, and is based on the Canton/Guangzhou dialect. Go'on 呉音 and kan'on 漢音 are both on'yomi/borrowed phonetic readings of Chinese characters, go'on being the earlier version.


I don't know Japanese, but, coincidentally, I have just saw this page:

Learning the most common phonetic and semantic chunks (or “radicals”) enables you to make educated guesses about the pronunciation and meaning of new characters. For example, all of the following Kanji share the same phonetic chunk, 工 (“craft”). It is pronounced kou (こう), and low and behold, each of the following Kanji it contains are all pronounced kou: 紅 (“crimson”)
空 (“empty”)
虹 (“rainbow”)
江 (“creek”)
攻 (“aggression”)
功 (“achievement”)

In the《詩經》(Canon of Poetry) era, phono-semetic characters (形聲字) always sounded same as the "sound" part. In fact, this is how phono-semetic convention works that time: when a character is not yet formed, we borrow the homophone and add another part indicate its meaning.

This is the 白一平 construction I looked up in 東方語言學網:

空: khoŋ
虹: kroŋs
江: kroŋ
攻: koŋ
功: koŋ

But after 3000 years, it ends up in Modern Mandarin(北京話) that the [r] and [s] are lost, [kh] becomes the fricative [h], or hardens in non-aspirated [k], or the [ʨ] the voiceless alveolo-palatal affricate, and they all develop different tones. They are still vaguely similar in position, but also widely different.

So, as of what you are thinking, my suggestion is to forget it, and just memorize it. There is no rule in general, except vague likeness. If you read extensively in Ancient Chinese (上古漢語) journals you might learn some arcane rules scholars established, but then you are spending several times long to learn those words, than just memorizing them! ;)


When I learned Cantonese, I started by first learning the pronunciation and romanization, then I memorized the Cantonese pronunciation of most common 3000 characters based on Jun Da's Modern Chinese Character Frequency List, since I already knew the characters, it was a breeze to learn the pronunciations.

One of the very interesting things that happened shortly after I began memorizing was that I began to see the relationships between the Mandarin Pronunciation and Cantonese and often have a very accurate gut-feeling for what the Cantonese pronunciation would be. I would imagine this may happen for you as well, and you will certainly very quickly begin to see the clues from the radicals based on other characters that you have memorized, but even i that weren't the case, 3000 characters isn't that bad when you're just learning pronunciation and then you can go on to the next step.

After I memorized about 1500 characters, I would read aloud for 20 minutes each day, this worked wonderfully since the writing is the same (although i had to be on the look-out for differences when it came to speaking).

As I was doing that, I would force myself to try to use Cantonese every chance I got. This made for some good laughs but with the base i had built from reading, it was very easy to transition into speaking.

Your journey will be a bit different in that if you do go through and memorize the pronunciations: 1) you will come across some new characters (which you will presumably want to learn along the way); 2) having memorized the pronunciations, you will not necessarily be able to understand what you are reading, but I would suggest that reading out loud for a short period each day will be very helpful regardless of whether you understand that much.

If you decide to tailor this approach to your needs, you may find the following anki deck helpful (after you swap the questions and answers so that it is prompting you with the character and you have to come up with the pronunciation): https://ankiweb.net/shared/info/39888802

Best of luck and let us know what solution you come up with :)

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

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