The Japanese Writing System
Modern Japanese uses a combination of:
(1) Chinese characters, known in Japanese as kanji ().
There are currently 1,945 officially approved characters for use in the media,
etc, although quite a few more are actually in use.
(2) Two kana syllabaries, hiragana and katakana, which
are 'alphabets' based on syllables rather than single sounds. Each syllabary
has a total of 46 basic letters, more if modified letters (letters with voicing,
reduced-size letters, etc.) are included.
(3) Roman letters and Arabic numbers borrowed from the West.
Chinese books were first brought to Japan between the 3rd and 5th centuries
A.D. Subsequently, the Japanese borrowed the Chinese writing system in its
entirety by adopting Classical Chinese as the official written language. It
was not long, however, before the Chinese script (with some adaptive modifications)
was being used to write Japanese. The Kojiki
(Record of Ancient Matters) in A.D. 712 wrote Japanese words and endings with
Chinese characters, mainly by using characters
for their phonetic value. This early system was a rather clumsy workaround
and eventually certain characters were simplified for specifically
phonetic purposes, resulting in the kana syllabaries.
The writing system underwent more than a few changes over the next thousand
odd years, finally reaching something like its present form following a 19th
century movement to 'unify speech and writing'. This system was subject to further
simplification and reform after the Second World War.
In the modern language, Chinese characters (kanji) are used to write
the major content words -- words with semantic content such as verbs, nouns,
and adjectives. Hiragana is used to write grammatical markers and endings.
Foreign words are written in katakana.
However, the system is rather more complex than this suggests. The best way
to grasp the Japanese writing system is to see it as the outcome of a struggle
to adapt Chinese characters to the writing of a completely different language,
involving many makeshift strategies and compromises.
The following is a sample of Japanese writing, a news story about an attempt
to extort money out of a former national volleyball coach. Note the use of
a mixture of Chinese characters, hiragana, and katakana. Numerals are international
style (Arabic numerals).
CHINESE CHARACTERS (KANJI)
An obvious and frequently-asked question is: Are the characters used in Japan
the same as those used in China?
The short answer is yes. Chinese characters used by the Japanese derive from
the same tradition as the Chinese, namely, the characters of the Classical language.
In the same way that almost any English word is fair game for adoption into
Japanese, virtually the entire corpus of traditional Chinese characters can
theoretically be used.
However, anyone familiar with the two languages quickly realises that there
are differences in the characters used. The main ones are:
- Number of characters in use: First, Japanese uses fewer characters
than Chinese. While the Japanese adopted much Chinese vocabulary, they did
not adopt everything. When they decided to restrict the number of kanji in use after the war, they were able to settle on 1,850 characters (since
increased to 1,945) without undue problems. Even though many writers and
academics find this number too restrictive and use some non-approved characters,
the total number still does not come near the 3,000-4,000 minimum
needed to function in Chinese. Japanese are aided by the fact that they
can fall back on hiragana or katakana to write their language,
unlike the Chinese who virtually use only characters.
- Non-classical characters in Chinese: Chinese contains some characters
that have been specifically created for post-Classical Chinese words. For
'look for' and ling4
'other' are relatively speaking fairly modern characters used to represent
post-Classical vocabulary. They are not normally used in Japanese.
- Japanese characters (kokuji): Where the Japanese could not
find an appropriate Chinese character to represent a Japanese word, they often
created their own. Such characters are known as kokuji ('national characters')
and a few have been included in the list of approved characters. Interestingly,
the Chinese regard kokuji as part of the greater family of Chinese
characters and include them in the larger, more comprehensive Chinese character
dictionaries. Some examples:
komu 'to be packed'
'to put into, be included', 'to load', 'concentrate on'
tsuji 'crossroads', personal name
- Standardisation of alternative forms: Many Chinese characters have
several alternative forms. The Chinese and Japanese sometimes decided on
different standard forms, e.g., the
character for 'receive', which is in
Chinese and in Japanese
(the Chinese contains an extra stroke).
- Simplification: The simplification of characters after WWII in Japan
and in the 1950s in China gave rise to divergences between the traditional
characters (still used in Hong Kong and Taiwan), China's simplified characters,
and Japan's simplified kanji. A Japanese character will sometimes be
the same as the traditional form, sometimes the same as the Mainland form,
sometimes the same as both, and sometimes different from both (see table below).
In general, the simplification of characters in Japan was relatively mild
compared to the drastic simplification on the Mainland.
Substitution of characters: In simplifying the characters the Japanese
also restricted the number to be used, disallowing the use of some characters
in education and the media. In order to follow these restrictions, abolished
characters are either written in hiragana or replaced with a different
character of similar meaning. This results in some less-than-satisfactory
||Simplified the same way in Mainland China and Japan
||Simplified in Mainland China but not Japan
||Simplified in Japan but not Mainland China (uncommon)
||'to transmit, story'
||Simplified differently in Mainland China and Japan
||Substitution of hiragana
||Substitution of different character
In such cases, the abolished character is present in spirit but not in
form, like a 'ghost limb' that has been amputated but feels as though
it is still there. Since hiragana are normally used to write verb
endings, substituting hiragana for characters can be confusing.
looks as though it should be read kusaran,
a literary or dialect pronunciation meaning 'not rot' -- the exact opposite
of furan! Moreover, replacing the disallowed
'soft, rot, fester, messy' with the approved character
ran 'be confused, in chaos' results in some
loss of meaning. 'Rotten and festering' is more accurate than 'rotten
Differences in the meanings of characters: Partly because of changes
in meaning on both sides, some characters
have different meanings in Chinese and Japanese. To take a simple example,
meant 'to run' in Classical Chinese. It still does in Japanese as
hashiru, but the meaning has changed to 'to
go' in modern Mandarin. From the Chinese point of view, the Japanese usage
has a quaint Classical feel to it.
the Japanese borrowed Chinese characters, the aspect that was to have the most
far-reaching consequence was the fact that they were used for both:
This results in the most striking feature of Chinese characters in Japanese,
i.e., the fact that most of them can be read in at least two ways,
one known as the on reading,
the other as the kun reading. This is a key concept in
understanding the structure of the modern Japanese vocabulary.
- Words (or forms) borrowed from Chinese -- the character was borrowed
with its pronunciation and meaning intact.
- Native Japanese words -- characters came to be identified with the native word having
the closest meaning.
The on reading is fairly easy to grasp: it's used for that large segment
of the Japanese vocabulary that was originally borrowed from Chinese.
The on reading thus approximates to the original Chinese pronunciation,
e.g., the character 'fire'
has the on reading ka, based on
the Chinese pronunciation of the time (the modern Mandarin pronunciation
'mountain' has the on reading san, zan
or sen, pronounced shan1
in modern Mandarin Chinese. Of course, the on readings are now rather
different from the modern Chinese pronunciation. This is due to various reasons
-- borrowing from different dialects, adaptation of the pronunciation to fit
the Japanese language, and changes on both sides over the past millennium.
But the similarities are quite obvious once allowances are made, and linguists
have even used the Japanese pronunciation as an aid in reconstructing the ancient
the on readings
roughly follow the Chinese pronunciation, the form of many characters gives
a hint of the pronunciation, just as they do in Chinese. For instance, the
following characters that have as
a phonetic all have similar on readings:
||'press, urge, force'
Multiple on readings:
One feature of the on readings is that many characters have more
than one possible on reading. For instance, the character 'sun,
day' (ri4 in
Chinese) has two on readings: nichi and jitsu.
It is read nichi in ichi-nichi 'one
day' and jitsu in honjitsu 'this
day'. This derives from the fact that the
Chinese writing system was borrowed not once but at least twice
during Japan's history. The first time resulted in what are known as the Go readings,
particularly associated with Buddhism (generally pre-Tang) and the second
was more secular in inspiration, resulting in the Kan readings (Tang
dynasty). Other, less important readings also exist
but these are isolated borrowings and do not form a coherent system like the Go and Kan readings.
For an example of such a reading, see gyooza.
The following table shows some examples of Go and Kan readings.
'sutra, pass through'
The Kan readings were eventually adopted as the officially correct
version and were supposed to supersede the Go readings, but it didn't
quite happen that way. Many Go readings continued in use,
especially in Buddhist contexts. Confusing as the existence of two
readings may be, the day might yet have been saved if only the language had
kept the two systems separate, i.e., by constructing individual words solely
or solely with Kan readings. Again, it didn't happen that way:
there are plenty of words in Japanese where one character is read in Go and the second character
character is read in Kan. Needless to
say, figuring out the correct reading can sometimes be a problem!
As in Chinese, characters form 'compound words', e.g.,
kazan 'fire' + 'mountain' = 'volcano'. In fact,
the on readings have taken on a life of their own, like Latin roots
in English. Many Chinese-style character compounds were actually
created by the Japanese in the modern era to translate Western terminology.
They have since been re-exported to China, resulting in a huge corpus of vocabulary
that is shared by Chinese and Japanese. This vocabulary can be read by speakers
of both languages, although with different pronunciations. This is a great
bond that links the two languages, despite the fact that they belong to entirely
different language families.
What truly complicates the Japanese writing system is the use of Chinese characters
to write native Japanese words. This custom originally arose as a makeshift
way of reading and understanding Chinese texts. Each character was glossed
with the Japanese word closest in meaning. Eventually a system grew up whereby
Chinese characters were assigned specific Japanese readings, known as kun
readings. For instance, it became customary for the character
('fire') to be read hi, which is the Japanese word
for 'fire'. Similarly,
('mountain') came to be read yama, the native Japanese
word for 'mountain'. Most (but not all) characters now have a kun reading
in addition to an on reading.
The assignment of Japanese readings to Chinese characters gave rise to several
1) Phonetic element is useless: Because the kun reading is native
Japanese, the phonetic element of the character is useless for guessing the
pronunciation. To take the earlier example using the phonetic element ,
the kun pronunciations are completely unpredictable, unlike the regular
pattern of the on readings:
||'press, urge, force'
2) Lack of fit between vocabularies: Since there is not a perfect fit
between the Chinese and Japanese vocabularies, complications arose in assigning
In some cases a single Japanese word corresponded to several different Chinese
words. For instance, the Japanese verb noboru meaning
'climb' or 'go up', corresponded to three Chinese words, as shown in this table:
||'go up' (in general)
|| 'rise (into the sky)'
||'climb (a mountain)'
The Japanese word noboru is now written with
three different characters. For the general meaning 'go up', it's written
(note: is hiragana
indicating the verbal ending -ru). In the meaning
'rise (into the sky)', it's written .
In the meaning 'climb (a mountain)', it's written .
This system generally works quite nicely, but sometimes even the Japanese
find it difficult to figure out which character is most appropriate (The
confusion that this can engender is detailed in Jack Halpern's page on what
he quaintly calls 'Japanese
The reverse also occurred. Where a single Chinese word corresponded to more
than one Japanese word, the same character was used for several different Japanese
words. For instance, the Chinese word
('above, on top of, go up') corresponded in meaning to ue
meaning 'above' or 'on top of', kami meaning 'above'
'higher', agaru meaning to 'go up', ageru
meaning 'to give' (respectfully), and noboru 'to
go up'. As a result,
is now used to write all these words.
||Japanese writing (multiple kun readings)
|| 'above' or 'on top of'
|| 'above' 'higher'
| 'to go up'
|| 'to give' (respectfully)
|| 'to go up, climb'
In the case of verbs with multiple readings, hiragana sometimes provides
a clue to pronunciation, e.g., agaru is written
+ -garu) and ageru
is written (
The divide in the Japanese vocabulary
The division between on and kun readings represents an important
divide in the Japanese vocabulary. Since words read with kun readings
are native Japanese words, they tend to be related to everyday life. (Alternatively,
they may be associated with older, more poetic Japanese words). On
readings are generally used for more formal, official, or scientific vocabulary.
The difference is similar to that found in English between everyday words
such as 'fire', and words using Latin or Greek roots such as 'ignite' (Latin ignis
= 'fire') or 'pyrotechnics' (Greek pur = 'fire').
Taking the character
'fire' as an example, the kun reading hi
is used in words with humble backgrounds like:
||'fire' + 'pot' = 'brazier'
||'fire' + 'seed' = 'live coal'
||'fire' + 'flower' = 'sparks'
||'down' + 'fire' = 'dying down'
The on reading ka is generally found
in more technical, formal, academic, or official words like:
|| 'fire' + 'power' = 'thermal power'
||'initiate' + 'fire' = 'ignition'
||'release' + 'fire' = 'arson'
||'fire' + 'funeral' = 'cremation'
||'war' + 'fire' = 'the fires of war'
||'come out' + 'fire' = 'break out'
On readings usually form a part of compound words as only a relatively
small number can stand as single words in Japanese.
Once the principle of on readings
and kun readings
is understood, the system falls into place and it is much easier to decide
which reading is required. However, there are still many complications that
must be overcome in order to read Japanese.
Despite the neat division between the two segments of the vocabulary, in practice
it has been impossible to keep them completely apart. There has been some
interesting crossover between on readings
and kun readings.
a) Words that have been converted from kun to on
shukka looks like a Chinese word. In
fact it was created as a fancy form of the ordinary Japanese expression
hi ga deru ('a fire breaks out'). In Chinese,
means 'to get angry'.
is another word that is not found at all in Chinese. It was created from the
hara ga tatsu (literally, 'abdomen stands up')
'to get angry'.
Some words that were originally native Japanese have also been converted
completely to on readings, e.g.
henji, which was originally the Japanese word
kaerigoto, a word that no longer exists. The
compound is not found
b) Mixtures of on and kun
Sometimes the crossover between
the two systems gives rise to mixed readings:
||Reading of First Part
||Reading of Second Part
'tiers of lacquered boxes'
|juu (on reading)
'place one on top of the other
|hako (kun reading)
|te (kun reading)
|hon (on reading)
Some words can be read with either a kun or an on reading,
e.g., 'town' in
place names can be read either choo (on reading)
or machi (kun reading), depending on
the locality. is
known as Honmachi in Osaka and Honchoo in
('water' + 'bird' = 'waterfowl') is normally read with the kun reading
mizu-dori but can also be read with the on
reading suichoo. In many cases it is simply
a matter of knowing the word, knowing how it's pronounced, and knowing how
it's written -- pretty much like learning English spelling.
There are several practices in the writing system that complicate matters even
- Chinese character compounds -- not just single characters -- are sometimes
used to write native Japanese words, especially the names of plants and
(purple + sun/yang
+ flower) is a Chinese word meaning 'hydrangea'. The hydrangea is known
in Japanese. To write ajisai,
Japanese takes over the compound ,
assigning the reading ajisai to the entire
compound without reference to the individual components (This is often
found in the writing of bird names -- see The
Writing of Japanese Bird Names Using Chinese Characters).
Similarly, the compound
meaning 'popular/widespread', can be used to write
the purely Japanese verb hayaru
'to be popular, widespread' -- not forgetting the hiragana verb ending:
- Various irregular readings have developed over the centuries and must
be learnt individually, e.g.,
('one day') is read tsuitachi when it means
'first of the month' (usually written across the page as ).
'lack + extend' is
quite irregularly read as akubi 'to yawn'. Place
names and personal names are a particularly rich source of idiosyncratic
readings (for instance, see Hard-to-read
place names in the Kansai region).
The result of the way that the Japanese have borrowed and applied Chinese
characters to their own language has given rise to considerable complexity
and a heavy burden for the learner of Japanese.
After all this complexity, the saving grace of Japanese, at least in comparison with Chinese, is that it has a way of writing the pronunciation of words, the kana syllabaries. As luck would have it, the Japanese have decided to use not one, but two different syllabaries to write their language.
HIRAGANA AND KATAKANA
The kana (hiragana and katakana) first developed because
Chinese characters could not properly represent the grammatical elements and
verb endings of Japanese. Originally certain Chinese characters were used
to write these elements phonetically. These characters were simplified over
time to yield the hiragana and katakana syllabaries. As the name
'syllabary' suggests, the syllabaries are used to represent combinations of
consonants and vowels, not single sounds. Except for certain details such as
the method of lengthening vowels, hiragana and katakana are completely
At the outset, the kana and the pronunciation of the language were fairly much in sync, but they inevitably drifted apart over time. By the 19th century, kana usage had become
confused due to changes in pronunciation.
A concerted attempt was made to reconstruct, restore, and enforce the original
usage of ancient times, which was called the 'historical kana usage'.
While this may have been correct in historical terms it was divorced from the modern pronunciation and placed a burden
on modern-day learners, resulting in calls
for reform. After World War II, kana usage
was reformed to reflect the modern pronunciation. Two redundant letters,
in their hiragana form, were abolished.
Hiragana: The hiragana script (written
in hiragana) is a cursive script still mostly used for:
Katakana: The katakana script (written
in katakana) is a blocky-looking syllabary now mainly used for
- Foreign words, e.g.,
depaato 'department store',
kurejitto kaado 'credit card',
kontenaa '(freight) container', etc. In the past
few decades, Japanese has subjected itself to a massive influx of foreign
words, resulting in a profusion of katakana.
- Scientific names of plants and animals, e.g.,
tobi 'kite', even
hito 'person' = 'homo sapiens'. All these words
are usually written in Chinese characters for non-scientific use.
- Onomatopoeic words, e.g.,
gata gata 'rattling/rickety sound'. (These words are similar to English
'thud', 'splash', and 'biff', but have a much wider range of application in
Japanese than English).
Furigana/rubi: With the heavy use of
characters in the past, it was common for printed texts to indicate the pronunciation
in small hiragana or katakana above the Chinese character, e.g.,
This practice, known as furigana or rubi, is no longer so
common but is still found in comics, books written for children, and academic
or other texts using difficult characters.
The full table of hiragana and katakana symbols can be found
at Hiragana & Katakana. More
hiragana and katakana tables can be found in the Links.
Could Kanji be replaced with Kana?
With such a complicated system of Chinese characters, the question naturally
arises: Can't the two kana syllabaries, hiragana and katakana,
be used in place of the characters? Theoretically the answer is yes, but in
fact the syllabaries still play only a supplementary role in the Japanese
script, much as they have from the start.
Aside from the weight of tradition and custom, there are several good reasons
why hiragana have failed to replace the characters.
1) Economising on space: The kana take up considerably more
space than characters. Nihon Keizai Shinbun ('Japan
Economic Newspaper', one of Japan's major newspapers) is
in characters and
2) Visual aid in reading: Characters create a focus when reading
text. Usually the main content words are in characters and the endings
are in kana.
Since Japanese does not put spaces between words, this is highly useful
in scanning text. A short phrase such as Tookyoo made
'as far as Tokyo' is easier to analyse into 'noun + particle' when characters
are used: . In this
provides a visual focus while the particle
made represents the attached grammatical element.
This effect is lost if hiragana only are used: .
3) Conveying a meaning: Characters suggest a meaning to the reader.
This not only helps give meaning to a text (for example, the characters
Tookyoo mean 'eastern capital'), it also helps
disambiguate words with similar pronunciations. For example,
hooka has a number of meanings, which become
clear when written in characters:
'arson', 'be released
from class', 'cannon
4) Link between on and kun vocabulary: Characters
act as a bridge between vocabulary using on and kun readings.
For instance, kuruma
'car' and jidoosha
'automobile' are completely unrelated words in hiragana, but when
written in characters as
the character (kuruma
in kun reading, sha in on reading,
meaning 'car') shows the connection between the two.
Similarly, characters may also act as a bridge among different on readings.
For instance, Keihin is a word used in expressions like 'the Keihin district' or 'the Keihin expressway'. It is an abbreviation of 'Tokyo-Yokohama'. This is not at all apparent until Keihin is seen in characters as .
It then becomes immediately clear that:
is from Tookyoo. This features a switch between two different on readings -- from the Go reading kyoo to
the Kan reading kei.
Yokohama. This features a switch from
the kun reading hama to the on reading hin.
Flexibility in Writing
The makeshift nature of the Japanese writing system means that the same word
may be written in different ways. While this can be a source of great confusion,
it can also offer writers a unique stylistic choice. The same word may be
written in either Chinese characters, hiragana, katakana, or
Roman letters depending on the situation, the type of text, and the writer's
preference. A few examples of the variation that can be found:
- In many everyday structural words, there is a choice between hiragana
and Chinese characters. For example,
aru 'to have, to exist' can be written
or (both pronounced
aru, but meaning 'to have' and 'to exist' respectively).
The word kono
'this' can be written
kono. Hiragana is modern standard usage,
but with word processors, which make it easy to summon up difficult characters,
the character versions are making something of a comeback. Men are more
likely to use characters in such cases. Heavy use of hiragana traditionally
tends to be identified with women's writing.
- The use of hiragana after verbs and associated nouns can show some
variation in usage. The verb mitsumoru 'to give
a cost estimate' is usually written .
The noun mitsumori, meaning 'cost quotation',
may be written ,
, or ,
the latter being preferred for its brevity in the word
mitsumori-sho '(written) cost quotation'. Correct
usage of hiragana after verbs has been subject to periodic changes
by the language and educational authorities.
- The word koohii 'coffee' is normally written
in katakana as
koohii, but where an old-fashioned, genteel
feeling is desired the old character rendition
koohii is sometimes used. Of course, the English
spelling 'coffee' is also quite common in advertising and Japanese have
no hesitation in reading it as koohii (see variants
in signs and advertising for specific examples in action).
- The word fakkusu may be written in katakana
as or using English
spelling as . The
latter appears more business-like and international.
- Sometimes katakana will be preferred to hiragana or characters
in order to set a word off from the rest of the sentence or because the
characters involved are awkward, unclear, or uncommon. E.g.,
boke 'idiot, stooge in stage act',
'bald', and ()
tsubo 'pressure point/acupuncture point'. This
frequently occurs with short two-syllable words, which appear 'punchier' written
- Choosing different characters can convey different meanings. For instance,
the Japanese word aoi meaning 'blue/green' can
be written with several Chinese characters, including ,
each indicating a different kind of 'blue' or 'green' - the blue of the sky,
the blue of the sea, the green of young leaves, the 'green' of a pallid face,
etc. A Japanese author can indicate exactly what kind of 'blue/green' he or
she is talking about by using a different character for aoi.
- A more interesting trick is to write a word in Chinese characters but indicate
with small lettering above that it should be read in a completely different
way. For instance, in translating the Harry Potter books, (see Harry
Potter in CJV), Yuko Matsuoka takes the character
ryuu or tatsu, meaning
'dragon' in the Oriental sense, and gives it the English reading 'dragon'
by merely adding the katakana for doragon
above the character: .
The beauty of this device is that familiar characters can be given completely
different readings, i.e., it's possible to write one thing and mean another.
It's also possible to use unfamiliar foreign words and explain their meaning
in Chinese characters.
An idea of the flexibility and variation that is possible with the Japanese
writing system can be seen in this look at signs and advertising
For more information on the Japanese writing system, see Links.
See also the Chinese Writing System and the Vietnamese