Sino-Japanese vocabulary, also known as kango (Japanese: 漢語, pronounced[kaŋɡo], "Han words"), is a subset of Japanese vocabulary that originated in Chinese or was created from elements borrowed from Chinese. Some grammatical structures and sentence patterns can also be identified as Sino-Japanese.
Kango is one of three broad categories into which the Japanese vocabulary is divided. The others are native Japanese vocabulary (yamato kotoba) and borrowings from other, mainly Western languages (gairaigo). It has been estimated that about 60% of the words contained in modern Japanese dictionaries are kango,[1] and that about 18–20% of words used in common speech are kango.[a] The usage of such kango words also increases in formal or literary contexts, and in expressions of abstract or complex ideas.[2]
Kango, the use of Chinese-derived words in Japanese, is to be distinguished from kanbun, which is historical Literary Chinese written by Japanese in Japan. Both kango in modern Japanese and classical kanbun have Sino-xenic linguistic and phonetic elements also found in Korean and Vietnamese: that is, they are "Sino-foreign", meaning that they are not pure Chinese but have been mixed with the native languages of their respective nations. Such words invented in Japanese, often with novel meanings, are called wasei-kango. Many of them were created during the Meiji Restoration to translate non-Asian concepts and have been reborrowed into Chinese.
Kango is also to be distinguished from gairaigo of Chinese origin, namely words borrowed from modern Chinese dialects, some of which may be occasionally spelled with Chinese characters or kanji just like kango. For example, 北京 (Pekin, "Beijing") which was borrowed from a modern Chinese dialect, is not kango, whereas 北京 (Hokkyō, "Northern Capital", a name for Kyoto), which was created with Chinese elements, is kango.
Ancient China's enormous political and economic influence in the region had a deep effect on Japanese, Korean, Vietnamese and other Asian languages in East and Southeast Asia throughout history, in a manner somewhat similar to the preeminent position that Greek and Latin had in European history. For example, the Middle Chinese word for gunpowder, Chinese: 火藥 (IPA:[xwa˧˥jak]),[3] is rendered as hwayak in Korean, and as kayaku in Japanese. At the time of their first contact, the existing Japanese language had no writing system, while the Chinese had a written language and a great deal of academic and scientific information, providing new concepts along with Chinese words to express them. Chinese became the language of science, learning, religion and government. The earliest written language to be used in Japan was literary Chinese, which has come to be called kanbun in this context. The kanbun writing system essentially required every literate Japanese to be competent in written Chinese, although it is unlikely that many Japanese people were then fluent in spoken Chinese. Chinese pronunciation was approximated in words borrowed from Chinese into Japanese; this Sino-Japanese vocabulary is still an important component of the Japanese language, and may be compared to words of Latin or Greek origin in English.
Sino-Japanese words are almost exclusively nouns, of which many are verbal nouns or adjectival nouns, meaning that they can act as verbs or adjectives. Verbal nouns can be used as verbs by appending suru (する, "do") (e.g. benkyō suru (勉強する, do studying; study)), while an adjectival noun uses -na (〜な) instead of -no (〜の) (usual for nouns) when acting attributively.
In Japanese, verbs and adjectives (that is, inflecting adjectives) are closed classes, and despite the large number of borrowings from Chinese, virtually none of these became inflecting verbs or adjectives, instead being conjugated periphrastically as above.
In addition to the basic verbal noun + suru form, verbal nouns with a single-character root often experienced sound changes, such as -suru (〜する) → -zuru (〜ずる) → -jiru (〜じる), as in kinjiru (禁じる, forbid), and some cases where the stem underwent a sound change, as in tassuru (達する, reach), from tatsu (達).
Sino-Japanese and on'yomi
The term kango is usually identified with on'yomi (音読み, "sound reading"), a system of pronouncing Chinese characters in a way that at one point approximated the original Chinese. On'yomi is also known as the 'Sino-Japanese reading', and is opposed to kun'yomi (訓読み, "reading by meaning") under which Chinese characters are assigned to, and read as, native Japanese vocabulary.
However, there are cases where the distinction between on'yomi and kun'yomi does not correspond to etymological origin. Chinese characters created in Japan, called kokuji (国字), normally only have kun'yomi, but some kokuji do have on'yomi. One such character is 働 (as in 働くhataraku, "to work"), which was given the on'yomi dō (from the on'yomi of its phonetic component, 動) when used in compounds with other characters, e.g. in 労働rōdō ("labor"). Similarly, the character 腺 ("gland") has the on'yomi sen (from the on'yomi of its phonetic component, 泉sen "spring, fountain"), e.g. in 扁桃腺hentōsen "tonsils"; it was intentionally created as a kango and does not have a kun'yomi at all. Although not originating in Chinese, both of these are regarded as 'Sino-Japanese'.
By the same token, that a word is the kun'yomi of a kanji is not a guarantee that the word is native to Japanese. There are a few Japanese words that, although they appear to have originated in borrowings from Chinese, have such a long history in the Japanese language that they are regarded as native and are thus treated as kun'yomi, e.g., 馬uma "horse" and 梅ume. These words are not regarded as belonging to the Sino-Japanese vocabulary.
While much Sino-Japanese vocabulary was borrowed from Chinese, a considerable amount was created by the Japanese themselves as they coined new words using Sino-Japanese forms. These are known as wasei-kango (和製漢語, Japanese-created kango); compare to wasei-eigo (和製英語, Japanese-created English).
Another miscellaneous group of words were coined from Japanese phrases or crossed over from kun'yomi to on'yomi. Examples include henji (返事 meaning 'reply', from native 返り事kaerigoto 'reply'), rippuku (立腹 'become angry', based on 腹が立つhara ga tatsu, literally 'belly/abdomen stands up'), shukka (出火 'fire starts or breaks out', based on 火が出るhi ga deru), and ninja (忍者 from 忍びの者shinobi-no-mono meaning 'person of stealth'). In Chinese, the same combinations of characters are often meaningless or have a different meaning. Even a humble expression like gohan (ご飯 or 御飯 'cooked rice') is a pseudo-kango and not found in Chinese. One interesting example that gives itself away as a Japanese coinage is kaisatsu-guchi (改札口 literally 'check ticket gate'), meaning the ticket barrier at a railway station.
More recently, the best-known example is the prolific numbers of kango coined during the Meiji era on the model of Classical Chinese to translate modern concepts imported from the West; when coined to translate a foreign term (rather than simply a new Japanese term), they are known as yakugo (訳語, translated word, equivalent). Often they use corresponding morphemes to the original term, and thus qualify as calques. These terms include words for new technology, like 電話denwa ('telephone'), and words for Western cultural categories which the Sinosphere had no exact analogue of on account of partitioning the semantic fields in question differently, such as 科学kagaku ('science'), 社会shakai ('society'), and 哲学tetsugaku ('philosophy'). While many terms were coined afresh (such as 科学 and 哲学), many were repurposed classical Chinese compounds, whose meanings were tenuously similar to their western counterparts. Here are a few examples:
Despite resistance from some contemporary Chinese intellectuals, many wasei kango were "back-borrowed" into Chinese around the turn of the 20th century. Such words from that time are thoroughly assimilated into the Chinese lexicon, but translations of foreign concepts between the two languages now occur independently of each other.[4] These "back-borrowings" gave rise to Mandarin diànhuà (from denwa), kēxué (from kagaku), shèhuì (from shakai) and zhéxué (from tetsugaku). Since the sources for the wasei kango included ancient Chinese texts as well as contemporary English-Chinese dictionaries, some of the compounds—including 文化bunka ('culture', Mandarin wénhuà) and 革命kakumei ('revolution', Mandarin gémìng)—might have been independently coined by Chinese translators, had Japanese writers not coined them first.[5] A similar process of reborrowing occurred in the modern Greek language, which took back words like τηλεγράφημα telegrafíma ('telegram') that were coined in English from Greek roots.[6] Many of these words have also been borrowed into Korean and Vietnamese, forming (a modern Japanese) part of their Sino-Korean and Sino-Vietnamese vocabularies.
Alongside these translated terms, the foreign word may be directly borrowed as gairaigo. The resulting synonyms have varying use, usually with one or the other being more common. For example, 野球yakyū and ベースボールbēsubōru both translate as 'baseball', where the yakugo 野球 is more common. By contrast, 庭球teikyū and テニスtenisu both translate as 'tennis', where the gairaigo テニス is more common. Note that neither of these is a calque – they translate literally as 'field ball' and 'garden ball'. ('Base' is 塁rui, but 塁球ruikyū is an uncommon term for 'softball', which itself is normally ソフトボールsofutobōru).
Finally, quite a few words appear to be Sino-Japanese but are varied in origin, written with ateji (当て字)— kanji assigned without regard for etymology. In many cases, the characters were chosen only to indicate pronunciation. For example, sewa ('care, concern') is written 世話, using the on'yomi "se" + "wa" ('household/society' + 'talk'); although this word is not Sino-Japanese but a native Japanese word believed to derive from sewashii, meaning 'busy' or 'troublesome'; the written form 世話 is simply an attempt to assign plausible-looking characters pronounced "se" and "wa". Other ateji of this type include 面倒mendō ('face' + 'fall down' = 'bother, trouble') and 野暮yabo ('fields' + 'livelihood' = 'uncouth'). (The first gloss after each character roughly translates the kanji; the second is the meaning of the word in Japanese.)
Types of on'yomi
On'yomi were originally used in ondoku (音読 "sound reading"), the Japanese system for reading aloud texts in the Middle Chinese (MC) language. A huge number of loanwords entered the Japanese language from Middle Chinese, intermediated by these conventionalized pronunciations. There are different types of on'yomi for Sino-Japanese vocabulary, depending mainly on the time period of borrowing.
Go-on (呉音 "Wu sound") readings represent the first major wave of Chinese borrowing in the 5th and 6th centuries, coinciding with the introduction of Buddhism in Japan. It is not agreed whether Go-on pronunciations are clearly derived from a particular dialect of Middle Chinese. Buddhist teachings along with the Chinese language were largely imported through the Korean peninsula, and it is unclear to what extent this fact influenced the Go-on pronunciations. Certain genres of modern vocabulary largely use Go-on readings, especially words related to Buddhism and law.
Kan-on (漢音 "Han sound") readings were introduced in the 7th through 9th centuries during the Tang dynasty, and are based on the central Chang'an pronunciation of Middle Chinese. While there was a large-scale effort to replace Go-on readings with Kan-on readings when pronouncing Chinese texts in Japan, this effort did not extend to changing the pronunciation of borrowed words that were already used in Japanese.[7] Massive borrowing of Chinese loanwords continued during this period, and these new borrowings reflected the new Kan-on readings. Today, Kan-on readings are the most commonly encountered type of on'yomi.
Kan'yō-on (慣用音 "customary sound") readings are not considered to follow the regular patterns for adapting either Go-on or Kan-on readings, but are commonly encountered in existing Sino-Japanese words. In some cases, the Kan'yō-on reading is in fact a regular development of the original Go or Kan on'yomi in a particular environment. For example, 拉 (MC lop) has the Kan'yō-on reading /raQ/ (or /ra/) in all Sino-Japanese words, which is the regular development of earlier /rap(u)/ before a voiceless obstruent. A common irregularity for Kan'yō-on is an unexpected voicing value for an initial obstruent. For example, 斬 (MC tʂɛmX) is read in all Sino-Japanese words as /zaN/ rather than the expected Kan-on reading /saN/.
Tō-on/Sō-on (唐音 "Tang sound" or 宋音 "Song sound") readings were introduced mostly from the 12th century onward, during and after the Song dynasty. "Tang" was in this context used to mean "Chinese" (i.e. "real Chinese pronunciation"), with no intended connection to the earlier Tang Dynasty.[7] Due to their more recent borrowing, Tō-on readings are sometimes more recognizably similar to Modern Chinese pronunciations. There are far fewer Sino-Japanese loanwords with Tō-on readings compared to Go-on and Kan-on readings. Dictionaries do not attempt to provide a Tō-on reading for each kanji as many do for Go-on and Kan-on readings.
Go-on and Kan-on readings have a special status when compared with other on'yomi types. Arising initially out of the need to be able to read any Chinese text aloud using ondoku, there is a long-standing practice of providing a Go and Kan reading for every kanji, even those which have never actually been used in borrowed Sino-Japanese vocabulary. The readings which are not actually encountered in Sino-Japanese loanwords were largely codified in the Edo period through the philological study of Chinese rime tables.[8] These readings are given in many dictionaries, though for the less common kanji there is sometimes disagreement between sources.
Correspondences between Middle Chinese and on'yomi
All characters used to write Middle Chinese represented a single syllable in the spoken language, made up of an "initial" (a single onset consonant), and a rime (the remainder of the syllable). Originally, the on'yomi for kanji attempted to closely match the Middle Chinese pronunciation for each character, while guided by the possible sounds and structures of Japanese as spoken at the time. In fact a number of new word shapes entered the language to accommodate the large influx of Chinese borrowings. Subsequently, many sound changes took place in Japanese, affecting both borrowed and native vocabulary. As such, on'yomi now often bear little resemblance to their original Middle Chinese source, and are even less similar to the pronunciation of the same characters in modern Chinese languages, which have undergone many changes from Middle Chinese. For example, 兄 (MC xjwæŋ) had the Go-on pronunciation [kwjaũ] when it was first borrowed, which subsequently developed to [kjaũ], then [kjau], then [kjɔː], and finally modern Japanese /kyō/ [kjoː].
Onsets (initials)
The Early Middle Chinese (EMC) initials have the following regular correspondences in Go and Kan on'yomi.
Aspiration was contrastive in Middle Chinese, but voiceless obstruents were adapted to Go and Kan pronunciations in the same way regardless of aspiration. However, many Kan'yō on'yomi exist with voiced obstruents corresponding to Middle Chinese unaspirated (and sometimes aspirated) voiceless obstruents. For example, 軍 (MC kjun) 'army' has the prescribed Go/Kan reading kun, but Kan'yō gun is the only reading actually used in Japanese. There are multiple reasons for the changes from the earlier Go to the later Kan pronunciations. These borrowings were drawn both from different times and different regions of China, and furthermore the Go pronunciations were likely intermediated through Korean Buddhist monks. However, there is little to support the claim that Go-on pronunciations were at the time of their introduction "less accurate" than their later Kan-on counterparts. The discrepancies between the two on'yomi categories are largely due to changes that took place between Early and Late Middle Chinese. The Early Middle Chinese (EMC) voiced obstruents became breathy voiced in Late Middle Chinese, e.g. [b > pɦ]. EMC [ɲ] became [ɻ], later becoming [ʐ] in Northern Chinese dialects. In the Japanese of both time periods, the voiced obstruents were prenasalized as [mb, nd, ndz, ŋg], helping to explain why they correspond to Middle Chinese nasals in Kan on'yomi. The Japanese consonant [p] developed first to [f] or [ɸ], and more recently to /h/ (with allophones [h, ɸ, ç]). Older [p] remains modern Japanese /p/ after the special moras /N/ and /Q/, and as such all /h/-initial on'yomi have regular variants with /p/ in this environment, for example Kan-on 筆 /hitu/ 'brush' vs. 鉛筆 /eN.pitu/ 'pencil'.
Rimes (medials and finals)
Middle Chinese rimes or "finals" contained a vowel, optional glides before the vowel (sometimes called "medials"), and an optional coda consonant /j, w, m, n, ŋ, p, t, k/— schematically (j)(w)V(C). The precise phonetic realization of the MC vowels is debated, and the set of vowels possible before different coda consonants varies considerably. When borrowed into Japanese, the more complicated MC vowel system was adapted to fit the Japanese five vowel system with /i, e, a, o, u/. MC rimes could begin with a glide /w/, /j/, or both /jw/. The earliest Japanese on'yomi allow the following sequences containing glides:
All of the /Cy/ and /Cw(y)/ sequences were newly introduced by borrowing from Chinese, though some would later arise in native vocabulary.[9] By the advent of the "historical kana" spellings (13th century, lasting until 1946[10]), the "ancient" kana sequences with /CwyV/ had long before lost their /w/, those with /Cwi/ had become /Cui, ki, gi/, and /ye/ merged with /e/. Later, /w/ was lost everywhere except in the sequence /wa/ with no preceding consonant. The presence of these glides in on'yomi is in some cases not easily predictable, for example 約 (MC ʔjak) has the Go reading yaku, while 央 (MC ʔjaŋ) has the jōyō Go reading ō, with yō listed as an alternate (but unused) Go reading.
The tables below show the regular correspondences between MC rimes and Japanese on'yomi (Go and Kan readings). The rimes are given in the transcription systems of Bernhard Karlgren, Li Rong, and William Baxter (see Middle Chinese finals for more transcription systems). Examples are given using the MC reconstructions from Karlgren's Grammata Serica Recensa (GSR), with the rimes transcribed using Baxter's system (see Character List for Karlgren's GSR). Japanese on'yomi are given in a phonemic transcription (see Japanese phonology).
Different MC rimes were restricted to following only certain MC initial consonants. Furthermore, the identity of the initial consonant sometimes results in a different regular outcome for the Japanese on'yomi. For the purposes of determining the Japanese on'yomi, the following sets of consonants can be distinguished:
Developments after the Japanese consonants /r/ (from MC /l/) and /n/ (from MC /n, ɳ, ɲ/) are noted where relevant. The MC onset /y/ (like all palatal onsets) appears only with MC rimes beginning in /j/, and generally patterns in on'yomi with MC /ʔ/ before the same rimes, but sometimes there is a distinction, where /y/ patterns with S. Where one of these five categories (P, T, S, K, Ø) appears in parentheses in the tables below, it refers to the adaptation of the MC rime after these different sets of consonants. Five columns in each table mark whether the given MC rime is found after each of these onset categories. A bullet (•) indicates that Go and Kan on'yomi exist corresponding to the given MC rime after the given onsets. When (~) appears, it indicates that an MC character exists which is expected to provide a relevant Japanese on'yomi, but it either has no identified reading, has on'yomi which are not clearly distinguished as Go vs. Kan, or has multiple MC pronunciations which make it impossible to determine which MC rime the on'yomi correspond to.
While the correspondences between MC rimes and Japanese on'yomi are rather consistent, there exists considerably more irregularity than is represented in these tables. Exceptional pronunciations are often found even for officially recognized Go and Kan readings. Furthermore, many kanji have Kan'yō-on readings, which by definition do not follow the regular correspondences, but appear in established Sino-Japanese words. The illusion of regularity is bolstered by the fact that lexicographers generally provide Go and Kan readings for characters based on their expected outcome, even when these readings are not actually employed in any Japanese word. Out of necessity, many of the examples shown below are of this type. Readings in the jōyō kanji list are highlighted in blue.
Vowel-final rimes
These MC rimes have no consonant after the vowel.
MC rime transcription
Occurs after
Japanese On readings
Examples
Karlgr.
Li
Baxter
P
T
S
K
Ø
Go
Kan
Middle Chinese, Go-on, Kan-on
ɑ
ɑ
a
•
•
•
•
•
a
波 pa, ha, ha; 多 ta, ta, ta; 左 tsa(H), sa, sa; 歌 ka, ka, ka; 荷 ɣa, ga, ka; 阿 ʔa, a, a
武 mjuX, mu, bu; 無 mju, mu, bu; 膚 pju, hu, hu; 屢/屡 ljuH, ru, ru; 俁 ŋjuX, gu, gu; 芋 ɣjuH, u, u 須 sju, su, syu; 娵 tsju, su, syu; 聚 dzjuH, zu, syu; 雛 dʐju, zyu, su
Rimes ending in a palatal glide
These MC rimes are analyzed as having a palatal glide after the vowel, though not all of the rimes end in a phonetic [j] in all MC transcription systems. These mostly end up as Japanese ai, e, ē, i, or ui.
The MC rimes ending in a labial glide were for the most part borrowed as diphthongs in Japanese. These later monophthongized as long vowels, such that these MC rimes mostly correspond to modern Japanese ō, yō, ū, or yū.
MC coda /m/ was originally written in Japanese with the man'yōgana无, which came to stand for the nasal special mora /N/. The manyō'gana无 developed into the hiragana ん used to represent /N/. It is possible that 无 originally represented two distinct sounds, moraic /m/ and moraic /n/ (from MC coda /n/, see below), but they may have been pronounced identically in Sino-Japanese vocabulary from the start. Regardless, 无 would not have stood for /mu/ in these words (the Go-on reading), just as the precursors of hiragana つ represented /t/ and not /tu/ when adapting the MC coda /t/ (see below). Native /mu/ from this time (man'yōgana牟 or 武, among others) remains /mu/, developing to /N/ only under very specific circumstances, while the borrowed moraic /m/ always develops to /N/.
MC coda /ŋ/ was borrowed as a single Japanese phoneme which was realized as two nasalized offglides: [ĩ] after /e/, and [ũ] after /u, o, a/. The nasality of these glides was generally not represented in writing, but in some cases was indicated with the same diacritic mark that would become the dakuten used to mark prenasalized obstruents.[9] These glides then denasalized, and the resulting diphthongs later monophthongized as long vowels. As such, almost all characters with the MC coda /ŋ/ end in ō, yō, ē, ū, or yū in modern Japanese on'yomi.
MC coda /p/ was borrowed as Japanese /pu/ (likely pronounced as [βu] after a vowel at the time of borrowing[11]). Note that these original readings are identical to the readings for MC /m/-final rimes, but with ふ in place of ん. The phoneme /p/ eventually lenited to /h/ word-initially, but was lost between vowels (except Vpa > Vwa). The result was that all /pu/-final readings developed /Vu/ sequences, which later monophthongized. This same change is seen in native vocabulary, as in OJ ke1pu > ModJ kyō 'today'. As a result of this development, all characters with the MC coda /p/ have Go and Kan readings ending in ō, yō or yū in modern Japanese.
Originally, borrowed coda /p/ functioned just like coda /t, k/ (see below) in that the "epenthetic" vowel /u/ did not appear before a voiceless obstruent /h~p, t, s, k/ in the same word, resulting in readings with the obstruent special mora /Q/ in place of /pu/. This phenomenon can still be seen in a number of Japanese words, for example 十 /zipu/ > /zyū/ 'ten' vs. 十歳 /ziQ.sai/ [dʑissai] 'ten years old' (now usually /zyuQ.sai/ [dʑɯssai]). For 拉 (MC lop), the expected Kan reading /rapu > rō/ is not found in Sino-Japanese vocabulary, but only /raQ/ as in 拉致 /raQ.ti/ [ɾattɕi] 'abduction' (shortened in most words to /ra/). However, for many characters, the vowel-final readings have been extended to all environments. In some cases, the reading with /Q/ led to the analogical creation of a /tu/-final reading. Notably, for 立 (MC lip) the Kan'yō-on reading /ritu/ (from regular /riQ/) is overwhelmingly common in Sino-Japanese vocabulary.
The MC coda /t/ was borrowed as Japanese /t/. Characters ending in this consonant were at first consistently pronounced with no epenthetic vowel, with the kana つ serving double duty to represent /t/ and /tu/.[9] Note that these readings are identical to the readings for MC /n/-final rimes, but with つ/ち in place of ん. Later, an epenthetic vowel /u/ or /i/ was inserted after the consonant in most environments. Kan-on readings use /tu/ exclusively, while the earlier Go'on readings use both /ti/ and /tu/ unpredictably. For example, MC 跋bat is adapted as Go /batu/, while the homophonous MC 犮bat is listed in dictionaries as Go /bati/ (though it is not actually used in existing Japanese words). Often Go readings with /ti/ and /tu/ are listed for the same character, though in practice those with /tu/ are much more common. For example, 滅 has the Go readings /meti/ and /metu/, but only /metu/ is recognized as the jōyō reading, and this is the only Go reading found in existing Japanese words. In fact only nine characters have jōyō readings with /(C)Vti/, though these include the common characters 一 /iti/ 'one', 七 /siti/ 'seven', 八 /hati/ 'eight', and 日 /niti/ 'day'. Before a voiceless obstruent /h~p, t, s, k/ in the same word, the epenthetic vowel does not appear, and the /t/ functions as the obstruent special mora /Q/, forming a geminate with the following obstruent. For example, 日 /niti/ 'day' appears as /niQ/ in the word 日記 /niQ.ki/ [nikki] 'diary'.
MC coda /k/ was borrowed as Japanese /k/ with a following epenthetic /i/ (after /e/) or /u/ (after /a, o, u/). After /i/, the epenthetic vowel (/iki/ vs. /iku/) depends on the original Middle Chinese vowel. The readings for MC /k/-final rimes are very similar to the original readings for MC /ŋ/-final rimes with く/き in place of nasalized う/い, but in this case there are some differences. Just like with coda /t/, the epenthetic vowel is absent before a voiceless obstruent /h~p, t, s, k/ in the same word, and the /k/ functions as the obstruent special mora /Q/. For example, 学 /gaku/ 'study' appears as /gaQ/ in the word 学校 /gaQ.kō/ [gakkō] 'school'.
All MC roots were a single syllable, and due to the restrictions on possible MC syllable shapes, a limited set of readings (on'yomi) are possible for borrowed Sino-Japanese roots. Furthermore, due in large part to the many distinct MC sounds which were merged when borrowed into Japanese, some readings are extremely common across different kanji, while others are very rare. The below table gives the number of kanji with each possible jōyōon'yomi (not distinguishing between Go, Kan, Tō, and Kan'yō, and not including readings considered restricted or rare). A zero represents a reading which is attested in Sino-Japanese vocabulary, but uses a non-jōyō reading. Readings which are listed in dictionaries but which are merely hypothesized and do not appear in attested Japanese words are not considered.
Number of jōyō kanji with each possible jōyō on'yomi (Go, Kan, Tō, Kan'yō)
on'yomi example word
#
on'yomi example word
#
on'yomi example word
#
on'yomi example word
#
on'yomi example word
#
on'yomi example word
#
on'yomi example word
#
on'yomi example word
#
i 医 i
25
e 恵 e
3
a 唖 a
1
o 汚職 o.syoku
2
u 有 u
5
ya 冶 ya
3
yo 預金 yo.kin
5
yu 柚 yu
7
hi 比 hi
21
ha 派 -ha
5
ho 歩 -ho
6
hu 麩 hu
22
ti 地 ti
12
ta 他 ta
4
to 都 to
12
tu 都合 tu.gō
1
tya 茶屋 tya.ya
1
tyo 著 tyo
2
si 詩 si
50
se 世界 se.kai
2
sa 差 sa
11
so 祖 so
13
su 素 su
3
sya 舎 sya
13
syo 書 syo
9
syu 主 syu
13
ki 気 ki
39
ke 卦 ke
3
ka 火 ka
32
ko 戸 ko
21
ku 口 ku
7
kyo 居 kyo
9
bi 美 bi
6
ba 罵声 ba.sē
3
bo 墓 bo
7
bu 武 bu
8
da 打開 da.kai
7
do 土 do
5
zi 字 zi
20
ze 是 ze
1
za 座 -za
2
zu 図 zu
2
zya 蛇 zya
2
zyo 女 zyo
7
zyu 綬 zyu
7
gi 義 gi
11
ge 下 ge
3
ga 画 ga
8
go 語 go
12
gu 具 gu
3
gyo 御 gyo
3
mi 味 mi
3
ma 魔法 ma.hō
4
mo 茂林 mo.rin
2
mu 無 mu
6
ni 二 ni
3
na 那辺 na.hen
2
nyo 女人 nyo.nin
2
ri 理 ri
9
ra 裸裎 ra.tē
3
ro 炉 ro
5
ru 流 ru
1
ryo 旅 ryo
4
wa 話 -wa
2
ē 栄 ē
10
ō 王 ō
15
ai 愛 ai
4
yō 用 yō
22
yū 勇 yū
17
yui 遺 yui
1
hē 平 hē
13
hō 方 hō
24
hū 風 hū
2
hai 肺 hai
10
hyō 票 hyō
8
tē 体 tē
24
tō 灯 tō
38
tū 通 tū
2
tai 態 tai
22
tui 追加 tui.ka
4
tyō 丁 tyō
27
tyū 中 tyū
14
sē 聖 sē
33
sō 草 sō
38
sū 数 sū
3
sai 犀 sai
24
sui 水 sui
12
syō 小 syō
57
syū 週 syū
25
kē 計 kē
31
kō 幸 kō
64
kū 空 kū
1
kai 回 kai
24
kyō 京 kyō
28
kyū 九 kyū
22
bē 米国 bē.koku
1
bō 坊 bō
23
bai 貝 bai
8
byō 秒 byō
6
byū 謬説 byū.setu
0
dē 泥土 dē.do
1
dō 堂 dō
11
dai 代 dai
5
zē 税 zē
1
zō 象 zō
9
zai 材 zai
5
zui 髄 zui
2
zyō 城 zyō
21
zyū 十 zyū
12
gē 芸 gē
3
gō 業 gō
9
gū 偶 gū
4
gai 害 gai
11
gyō 行 gyō
6
gyū 牛 gyū
1
mē 名 mē
8
mō 毛 mō
7
mai 枚 -mai
6
myō 妙 myō
5
nē 寧日 nē.zitu
1
nō 農 nō
6
nai 内 nai
1
nyō 尿 nyō
1
nyū 入力 nyū.ryoku
3
rē 令 rē
12
rō 蝋 rō
10
rai 雷 rai
4
rui 類 rui
4
ryō 両 ryō
16
ryū 竜 ryū
7
wai 歪度 wai.do
1
in 引 in
13
en 円 en
17
an 暗 an
3
on 音 on
5
un 運 un
2
hin 品 hin
5
hen 変 hen
7
han 半 han
24
hon 本 hon
3
hun 分 hun
8
tin 珍 tin
6
ten 天 ten
9
tan 短 tan
14
ton 遁辞 ton.zi
3
sin 新 sin
29
sen 戦 sen
30
san 三 san
12
son 損 son
6
sun 寸 sun
1
syun 旬 syun
3
kin 金 kin
16
ken 剣 ken
32
kan 漢 kan
46
kon 魂 kon
13
kun 訓 kun
4
bin 敏 bin
4
ben 弁 ben
3
ban 晩 ban
8
bon 盆 bon
2
bun 文 bun
3
den 伝 den
4
dan 男 dan
9
don 鈍 don
3
zin 仁 zin
11
zen 全 zen
8
zan 残 zan
4
zon 存意 zon.i
1
zun 寸胴 zun.dō
0
zyun 順 zyun
11
gin 銀 gin
2
gen 言 gen
12
gan 丸 gan
10
gon 勤求 gon.gu
1
gun 軍 gun
3
min 眠 min
2
men 面 men
4
man 万 man
4
mon 門 mon
5
nin 人 nin
5
nen 年 nen
6
nan 南 nan
4
rin 林 -rin
7
ren 連 ren
5
ran 乱 ran
6
ron 論 ron
1
wan 湾 wan
2
itu 逸 itu
1
etu 悦 etu
4
atu 圧力 atu.ryoku
1
otu 乙 otu
utu 鬱 utu
1
hitu 筆 hitu
4
hetu 丿乀 hetu.hotu
0
hatu 発 hatu
2
hotu 発願 hotu.gan
1
hutu 仏人 hutu.zin
2
titu 帙 -titu
2
tetu 鉄 tetu
5
tatu 達人 tatu.zin
1
totu 凸 totu
2
situ 室 situ
9
setu 節 setu
11
satu 札 satu
9
sotu 卒 sotu
2
syutu 出 syutu
1
kitu 橘 kitu
3
ketu 決 ketu
7
katu 活 katu
9
kotu 骨 kotu
2
kutu 窟院 kutu.in
3
betu 別 betu
2
batu 罰 batu
5
botu 没 botu
2
butu 仏 butu
2
detu 捏造 detu.zō (rare reading)
0
datu 脱 datu-
2
zitu 実 zitu
2
zetu 舌炎 zetu.en
2
zatu 雑 zatu
1
zutu 術無し zutu.na.si (rare reading)
0
zyutu 述 zyutu
2
getu 月 getu
1
gatu 歹偏 gatu.hen
1
gotu 兀然 gotu.zen (rare)
0
mitu 密 mitu
2
metu 滅亡 metu.bō
1
matu 末 matu
2
motu 没薬 motu.yaku
1
netu 熱 netu
1
natu 捺印 natu.in
0
ritu 律 ritu
4
retu 列 retu
4
ratu 辣腕 ratu.wan
1
iti 一 iti
2
eti 越後 eti.go
0
hati 八 hati
2
(tati) 達 -tati (possibly native)
0
siti 七 siti
2
seti 刹那 seti.na
0
soti 帥 soti
0
kiti 吉 kiti
1
keti 血縁 keti.en
0
kati 褐 kati
0
beti 別 beti
0
bati 罰 bati
1
(dati) 達 -dati (possibly native)
0
ziti 実 ziti (rare reading)
0
zuti 術無し zuti.na.si (rare reading)
0
gati 月 gati
0
miti 蜜 miti (rare reading)
0
moti 勿論 moti.ron
0
niti 日 niti
1
neti 熱 neti (rare reading)
0
riti 律 riti
0
rati 埒 rati
0
iku 育児 iku.zi
1
aku 悪 aku
2
oku 屋 oku
4
yaku 益 yaku
6
yoku 欲 yoku
6
haku 白 haku
9
hoku 北西 hoku.sē
1
huku 服 huku
9
hyaku 百 hyaku
1
tiku 竹簡 tiku.kan
5
taku 宅 taku
7
toku 徳 toku
7
tyaku 着 tyaku
2
tyoku 直 tyoku
3
seku 齷齪 aku.seku
0
saku 作 saku
10
soku 束 soku
11
syaku 勺 syaku
5
syoku 食 syoku
10
syuku 宿 syuku
6
kiku 菊 kiku
1
kaku 角 kaku
18
koku 国 koku
8
kyaku 客 kyaku
3
kyoku 局 kyoku
3
baku 獏 baku
5
boku 僕 boku
7
buku 服 buku (archaic)
0
byaku 白毫 byaku.gō
1
daku 濁点 daku.ten
2
doku 毒 doku
3
ziku 軸 ziku
1
zaku 石榴 zaku.ro
0
zoku 族 zoku
5
zyaku 弱 zyaku
3
zyoku 辱 zyoku
1
zyuku 塾 zyuku
2
gaku 学 gaku
5
goku 極 goku
2
gyaku 逆 gyaku
2
gyoku 玉 gyoku
1
maku 幕
2
moku 木 moku
3
myaku 脈 myaku
1
niku 肉 niku
1
nyaku 蒟蒻 kon.nyaku
0
riku 陸 riku
1
raku 陸 raku
4
roku 六 roku
3
ryaku 略 ryaku
1
ryoku 力役 ryoku.eki
2
waku 惑 waku
1
iki 域 iki
1
eki 役 eki
6
hiki 疋 -hiki
0
heki 癖 heki
3
teki 敵 teki
6
siki 式 siki
4
seki 席 seki
16
keki 砉然 keki.zen
0
beki 冪 beki
0
deki 溺愛 deki.ai
1
ziki 直 ziki
1
geki 劇 geki
4
riki 力 riki
1
reki 鬲 reki
2
Due to the fact that most MC syllables had a coda, most Japanese on'yomi are bimoraic, containing either two syllables, a long vowel, or the moraic nasal /N/. These last two structures are extremely common in Sino-Japanese roots, but somewhat rare in native Japanese vocabulary. For these and other reasons, the phonological patterns of Sino-Japanese words and native Japanese words are markedly different, and it is very often possible to correctly guess the etymological origin of a word based solely on its shape.
Phonetic correspondences between Modern Chinese and on'yomi
Comparison with Mandarin
At first glance, the on'yomi of many Sino-Japanese words do not resemble the Modern Standard Chinese pronunciations at all. Firstly, the borrowings occurred in three main waves, with the resulting sounds identified as Go-on (呉音), Kan-on (漢音), and Tō-on (唐音); these were at different periods over several centuries, from different stages in Historical Chinese phonology, and thus source pronunciations differ substantially depending on time and place. Beyond this, there are two main reasons for the divergence between Modern Standard Chinese and Modern Standard Japanese pronunciations of cognate terms:
Most Sino-Japanese words were borrowed in the 5th–9th centuries AD, from Early Middle Chinese into Old Japanese. Both languages have changed significantly since then, and in different ways. This has resulted in the respective pronunciations becoming more and more divergent over time.
Middle Chinese had a much more complex syllable structure than Old Japanese, as well as many more vowel and consonant differences. Many sounds and sound combinations had to be approximated in the borrowing process, sometimes with significant differences (e.g. final /ŋ/ was represented as /u/ or /i/).
Nonetheless, the correspondences between the two are fairly regular. As a result, Sino-Japanese can be viewed as a (transformed) "snapshot" of an archaic period of the Chinese language, and as a result is very important for comparative linguists as it provides a large amount of evidence for the reconstruction of Middle Chinese.
The following is a rough guide to equivalencies between modern Chinese words and modern Sino-Japanese on'yomi readings.
Unless otherwise noted, in the list below, sounds shown in quotation marks or italics indicate the usage of non-IPA romanization such as Hanyu pinyin for Mandarin Chinese and Hepburn romanization for Japanese. Symbols shown within slashes or square brackets, like /ɡ/ or [dʒ], are IPA transcriptions.
A major sound-shift has occurred in Mandarin since the time of modern contact with the West. Namely, the sounds written in Pinyin as "g" [k] or "k" [kʰ], when immediately preceding an "i", "y" or "ü" sound, became "j" ([tɕ], similar to English "j") or "q" ([tɕʰ], similar to English "ch"). This change is called palatalization. As a result, Peking (北京) changed to Běijīng, and Chungking (重慶) to Chóngqìng. This shift did not occur in Sino-Japanese. Thus, Mandarin qì (氣, 'breath, air, spirit') corresponds to Japanese ki. In some other varieties of Chinese, it is still pronounced as 'ki'. For example, 氣 in Southern Min is khì (Pe̍h-ōe-jī romanization). This is similar to the way the Latin ⟨C⟩, once always pronounced like an English ⟨K⟩, became closer to an English ⟨CH⟩ in Italian words where the ⟨C⟩ is followed by an ⟨E⟩ or ⟨I⟩, changing centum/kentum/ into cento/tʃento/.
Old Japanese did not have an "-ng" or [ŋ] syllable ending, which is very common in Chinese. This sound was borrowed as either /i/ or /u/. The combinations /au/ and /eu/ later became "ō" and "yō", respectively, in Japanese. Thus, the Mandarin reading of "Tokyo" (東京; Eastern (東) Capital (京)) is Dōngjīng; this corresponds to Japanese Tōkyō, with sound history for 京 being supposed approximately *kiæŋ -> kyau -> kyō (for comparison: Southern Min 京 (colloquial) is kiaⁿ with a nasal diphthong). Another example is 京城, former name for Seoul, which is Keijō in Japanese and Gyeongseong in Korean (which did and does have syllables ending in [ŋ]). 京 is read "kei" (*kiæŋ -> kyei -> kei) in this case.
As in the case of 京, the same character sometimes has multiple readings, e.g. "kyō" (Go-on) vs. "kei" (Kan-on) vs. "kin" (Tō-on). These stem from multiple phases of borrowing, which occurred at different times and from different source dialects and were carried out by different groups of people possibly speaking different dialects of Japanese. This means that the same word may have had different Chinese pronunciations, and even if not, the borrowers may have chosen different strategies to handle unfamiliar sounds. For example, the character 京 seems to have had an approximate pronunciation of /kjæŋ/ at the time of both the Go-on (5th–6th century AD) and Kan-on (7th–9th century AD) borrowings; however, the unfamiliar vowel /æ/ was represented by /a/ in the former case and /e/ in the latter. (This may also indicate different source pronunciations of the vowel.) In addition, the unfamiliar final /ŋ/ was represented by /u/ in the former case but /i/ in the latter, agreeing in frontness vs. backness with the main vowel. By the time of the Tō-on borrowing (post-10th century), the pronunciation in Chinese had changed to /kiŋ/, thus the pronunciation "kin" was decided as the closest approximation.
The vowels of Chinese sometimes correspond to Sino-Japanese in an apparently haphazard fashion. However, Mandarin "ao" often corresponds to Japanese "ō" (usually derived from earlier Sino-Japanese [au]), and Chinese empty rime[ɨ] (represented in pinyin with a "i") often corresponds to [i] (a different sound, also represented with a "i" in Hepburn) in Japanese.
The distinction between voiced and unvoiced consonants ([d] vs. [t] or [b] vs. [p]) has been lost in modern Mandarin and many other varieties of Chinese. The key exception is in Wu dialects (呉語, e.g. Shanghainese). The Shanghainese voiced consonants match the Japanese go-on (呉音) readings nearly perfectly in terms of voicing. For example, 葡萄 (grape) is pronounced "budo" in Shanghainese and "budō" (< "budau") in Japanese (preserving the voiced consonants [b] and [d]), but "pútáo" in Mandarin. Incidentally, the rising tone of the Mandarin syllables may reflect the earlier voiced quality of the initial consonants.
In modern Mandarin, all syllables end either in a vowel or in one of a small number of consonant sounds: "n", "ng", or occasionally "r". However, Middle Chinese, like several modern Chinese dialects (e.g. Yue, Hakka, Min), allowed several other final consonants including [p], [t], [k], and [m], and these are preserved in Sino-Japanese (except for -m, which is replaced by -n, as in 三, san, "three"). However, because Japanese phonology does not allow these consonants to appear at the end of a syllable either, they are usually followed in Sino-Japanese by an additional "i" or "u" vowel, resulting in a second syllable (-tsu or -chi if from -t, -ku or -ki if from -k, and -pu if from -p, although -pu became -fu and then simply -u). As a result, a one-syllable word in Chinese can become two syllables in Sino-Japanese. For example, Mandarin tiě (铁, 'iron') corresponds to Japanese tetsu (鉄). This is still pronounced with a final [t] in Cantonese: /tʰiːt˧/ (Vietnamese thiết). Another example is Mandarin guó (國, 'land'), from Early Middle Chinese /kwək/, corresponding to Japanese koku.
The consonant "f" in Mandarin corresponds to both "h" and "b" in Japanese. Early Middle Chinese had no /f/, but instead had /pj/ or /bj/ (in other reconstructions, /pɥ/ or /bɥ/). Japanese still reflects this ("h" was /p/ in Old Japanese). For example, Mandarin Fó (佛 'Buddha') corresponds to Japanese butsu (仏); both reflect Early Middle Chinese /bjut/ from a still older form /but/. In modern Southern Min Chinese, this character may be pronounced either [put] or [hut] (colloquial and literary respectively).
In addition, as in the previous example, Old Japanese /p/ became modern "h". When a Middle Chinese word ended in /p/, this produced further complications in Japanese. For example, Middle Chinese /dʑip/ 'ten' (Standard Mandarin "shí", Cantonese /sɐp/) was borrowed as Old Japanese /zipu/. In time this went through a series of changes: /zipu/ > /zihu/ > /ziu/ > /zjuː/[12] > "jū". Note that in some compounds, the word was directly borrowed as /zip-/ > "jip-"; hence "jippun" 'ten minutes' (or "juppun", influenced by "jū"), rather than "*jūfun".
More complex is the archaic dento-labial nasal sound: The character 武 ('strife, martial arts') was pronounced "mvu" in Late Middle Chinese. The sound is approximated in the Japanese pronunciations "bu" and "mu". However, that sound no longer exists in most modern Chinese dialects, except Southern Min "bú", and the character 武 is pronounced "wǔ" in Mandarin, /mou˩˧/ in Cantonese, "vu" in Hakka, Shanghainese, and Vietnamese.
The modern Mandarin initial "r" usually corresponds to "ny" or "ni" in Japanese. At the time of borrowing, characters such as 人 ('person') and 日 ('day'), which have an initial "r" sound in modern Mandarin, began with a palatal nasal consonant [ɲ] closely approximating French and Italiangn and Spanishñ. (This distinction is still preserved in some Chinese varieties, such as Hakka and Shanghainese, as well as Vietnamese.) Thus Mandarin Rìběn (日本, Japan) corresponds to Japanese Nippon. This is also why the character 人, pronounced /ɲin/ in Middle Chinese, is pronounced "nin" in some contexts, as in ningen (人間), and "jin" in others, such as gaijin (外人)— approximating its more modern pronunciation. In Wu dialects, including Shanghainese, 人 ('person') and 二 ('two') are still pronounced "nin" and "ni", respectively. In Southern Min (especially Zhangzhou accent), 人 is "jîn" (literary pronunciation) which is practically identical to Japanese On'yomi.
In Middle Chinese, 五 ('five') and similar characters were pronounced with a velar nasal consonant, "ng" ([ŋ]), as its initial. This is no longer true in modern Mandarin, but it remains the case in other Chinese dialects such as Cantonese (/ŋ̩˩˧/) and Shanghainese. Japanese approximates the Middle Chinese */ng/ with "g" or "go"; thus 五 becomes "go". In Southern Min, it is pronounced /gɔ/ (colloquial) or /ŋɔ/ (literary) while in the Fuzhou dialect it is pronounced "ngu". In addition, some Japanese dialects have [ŋ] for medial g.
The Mandarin "hu" sound (as in "huá" or "huī") does not exist in Japanese and is usually omitted, whereas the Mandarin "l" sound becomes "r" in Japanese. Thus, Mandarin Huángbò (黄檗) corresponds to Japanese Ōbaku, and Rúlái (如来) and lamian (拉麵) to Nyorai and ramen respectively.
Mandarin "h", usually from Middle Chinese [x] or [ɣ] will often correspond to "k" or "g" in Japanese, as Old Japanese lacked velar fricatives: Modern Japanese [h] is derived from Old Japanese [ɸ], which descended in most cases from a Proto-Japonic */p/; however, this lack of velar fricatives in Old Japanese helps preserve the voiced-voiceless contrast between Middle Chinese [x] and [ɣ] that Mandarin, Cantonese, Korean and Vietnamese has lost. Mandarin "z" will often correspond to Japanese "j"; these are also changes in Chinese. Thus, Mandarin hànzì (漢字) corresponds to Japanese kanji, hànwén (漢文, Chinese written language) to kanbun, and zuìhòu (最後 'last') to saigo.
Chart of correspondences
Note:
MC: Middle Chinese
Pinyin: Modern Standard Chinese (Mandarin) in its official spelling. Multiple outcomes for MC initials (e.g. MC /ɡ/ > Pinyin g,j,k,q) are primarily due to two reasons:
MC voiced stops/affricates become Mandarin aspirated stops/affricates (p,t,k,etc.) when the syllable had the MC first tone (Mandarin first/second tones), unaspirated stops/affricates (b,d,g,etc.) otherwise.
Early Mandarin velar obstruents (g,k,h) and alveolar sibilants (z,c,s) become palatal obstruents (j,q,x) when a front vowel or glide followed.
^Baxter represents this MC initial with <y> to distinguish it from the palatal glide <j> which appears at the beginning of many rimes. Karlgren distinguishes them as <j> and <i̯>.
^/kwe, gwe/ were originally indicated in the historical spelling, but are not used in the version that became standard.
^Ancient spellings with /ye/ are not given in most dictionaries, and the tables below will not attempt to distinguish between on'yomi originally pronounced with /ye/ vs. /e/.
^There is at least one example of ancient /kwyo/: 洫 くゐよく /kwyoku/.
^The Dai Kan-Wa Jiten treats the kana for /twi, dwi, swi, zwi, rwi, ywi/ as historical spellings, but all other sources treat them only as ancient.
^Chung, Karen S. (2001). "Chapter 7: Some Returned Loans: Japanese Loanwords in Taiwan Mandarin". In McAuley, T. E (ed.). Language change in East Asia. Richmond, Surrey: Curzon. pp. 161–163. ISBN0700713778.
^Martin, Samuel Elmo (1987), The Japanese language through time, Yale University Press, cited in Cécile Fougeron; Barbara Kuehnert; Mariapaola Imperio; Nathalie Vallee (31 August 2010). Laboratory Phonology 10. Walter de Gruyter. p. 207. ISBN978-3-11-022491-7.