Home Previous Columns Book Reviews Other Articles Reader Response Links

Read an interview with James Heisig on how his method was born.

Remembering the Kanji I, II, and III
James W. Heisig
University of Hawaii Press, 2007, 2008

by Gary E. Harper (harper@ii-okinawa.ne.jp)

Many references have been made to James W. Heisig's three books, Remembering the Kanji, I, II and II at KanjiClinic.com. Recently I took the time to look again carefully at these books, particularly the author's comments. I would like to begin by making some general observations about Heisig’s system, which I have used in my own kanji study.

1.Dr. Heisig realizes that it is much easier to learn one thing at a time, and has incorporated this idea into his books. NO ONE else (to my knowledge) has done this with regards to kanji study for foreigners. All other books give the student 書き順、(stroke order), 音読み/訓読み(on/kun pronunciations), and the English translations, all in one indigestible lump.

2. Heisig also realizes that learning kanji is not equivalent to learning vocabulary. (The task of learning kanji seems to me to be closer to learning to spell, admittedly using an extremely complex system).

3. Heisig uses mnemonics to associate the meaning of the kanji with the English keyword and with the individual elements within the kanji, and while the use of mnemonics for learning kanji is not new--every elementary-school level kanji book includes mnemonics for the child to hold on to--he often uses mnemonics that are particularly easy to remember. Books for foreigners tend to use the derivation of the character, which may be less vivid for memory purposes.

4. After the first 508 characters, Heisig asks the student to make up his/her own mnemonics, still based on the elements of the characters (he calls them"primitives") which he supplies. Note that this cannot help but develop one's ability to create mnemonics, which is a very useful skill, not only for learning kanji but for learning other things as well.

5. Heisig insists that writing individual characters a number of times is a waste of time, time better spent drilling with flashcards.

6. In reviewing flashcards, a process Heisig strongly recommends, he insists you review ONLY from the English keyword to the kanji, and not the reverse. (Since reading is a strong objective, most of us-- I at least-- tend to review from the Japanese to the English). I have no idea why, but reviewing from the English to kanji works extremely well, just as Heisig claims it will.

Remembering the Kanji I, however, provides only the kanji and one English key word for each. Kanji with similar components are grouped together. If one learns, say, 20 kanji per day, it will take only about 100 days to learn the shapes and one English meaning for each of the 1,945 joyo (general-use) kanji using Heisig’s first volume.

The upshot of the above is that the student can learn the shapes of the joyo kanji in a short period of time, including stroke order. The student will have only a general idea of the meanings of many kanji, and will have no idea of the 音読み(on pronunciation) or 訓読み (kun pronunciation), but he/she will have a foundation that is close to that of a student from a kanji-using culture.


Remembering the Kanji II more or less completes the job. Here, Heisig provides lists of characters having the same 音読み based on the fact that they share a phonetic component. He provides ONLY ONE 音読み, recommending that the student learn others as he/she meets them in reading. He also provides ONE 熟語 (jukugo, compound word), usually a commonly used one, to provide a context for the 音読み. Since the student has likely been studying Japanese for some time at this stage, he/she may well already know the 熟語 involved, or may wish to use another one he/she does already know as a "handle" for the 音読み. The handle Heisig provides is, as far as possible, related to the English key word.

Many kanji, despite having a certain phonetic component, (which the author calls "signal primitives”), do not have the 音読み normally associated with it, and these “stragglers” are listed in separate chapters. The 701 characters with no phonetic component are also listed separately.

Since one 熟語 is given to place the 音読み in context, in the best case, two mnemonic devices are provided. Where there is no "signal primitive" the student must use the 熟語 as the mnemonic, but as this chapter comes after all the others, the student should be able to assimilate these kanji fairly easily by then .

For 訓読み (kun pronunciation), which the student is expected to tackle AFTER finishing 音読み、Heisig provides a rather ingenious mnemonic system involving the meanings inherent in each hiragana. I personally haven't yet used this, but it may be helpful for learning the kunpronunciation of words one does not often encounter in reading.

Both 音読み and 訓読み are meant to be transcribed to the flash cards used in Volume I, and reviewed from pronunciation to kanji, not the reverse. By the way, Heisig strongly recommends against the use of romaji in learning pronunciations.


Heisig suggests reading and using the kanji as much as possible. He further mentions that memorizing lists of vocabulary words is a waste of time and will result in not remembering most words. I must agree, but I have found that putting the vocabulary words I want to memorize into sentences greatly aids in fixing the word in my mind.

Heisig also notes the futility of trying to memorize 熟語 homonyms. I agree completely that memorizing 先生、先制、宣誓、陝西、and 専制 (all pronounced “sensei”) as a group is a waste of time, but when the meanings are close, or when the 熟語 are likely to be confused, I think the differences should be remembered (not memorized out of context, however). There are numerous low-cost books for native speakers of all education levels on 間違いやすい漢字 (easily confused kanji).

VOLUME III follows the same methods used in Vols. 1 and 2, and teaches (fairly) commonly used non-joyo kanji. It is meant for the advanced student who already knows the joyo kanji.

In the introduction to Volume III, Heisig states that, even if many students find the books useful, the method is essentially useless for classroom instruction. I must agree. I have visions of a test including questions like "What kanji is connected with two oysters in shell-to-shell combat?" But the method is very good for individual study.

Read a review of Volume 1.
Read a review of Volume 2.