The classical Japanese language (文語 bungo, "literary language"), also called "old writing" (古文 kobun), is the literary form of the Japanese language that was the standard until the early Shōwa period (1926–89). It is based on Early Middle Japanese, the language as spoken during the Heian period (794–1185), but exhibits some later influences. Its use started to decline during the late Meiji period (1868–1912) when novelists started writing their works in the spoken form. Eventually, the spoken style came into widespread use, including in major newspapers, but many official documents were still written in the old style. After the end of World War II because of the surrender of Japan, most documents switched to the spoken style, although the classical style continues to be used in traditional genres, such as haiku and waka. Old laws are also left in the classical style unless fully revised.
Contents
- Orthography
- Old character forms Kyjitai
- Historical kana usage Rekishi teki kana zukai
- H Row Ha gy rule
- W row Wa gy rule
- D row Da gy rule
- Y row Ya gy rule
- Geminate Sokuon rule
- Labialized consonant Gyon rule
- Classical auxiliary verb mu rule
- Miscellaneous
- Conjugation table
- Verb class distribution
- Adjectives Keiyoushi
- Adjectival verbs Keiyou doushi
- References
Orthography
Classical Japanese is written in an orthography that differs from modern Japanese in two major ways. These are the usage of old character forms (旧字体 Kyūjitai) and historical kana usage (歴史的仮名遣 Rekishi-teki kana-zukai).
Old character forms (旧字体 Kyūjitai)
Old character forms are the forms of Chinese characters (漢字 Kanji) used in Japan prior to the post-World War II spelling reforms in that country. The modern, simplified characters are called new character forms (新字体 Shinjitai).
A few examples follow, with the old characters on the left and the new characters on the right (pronunciations are the most common Japanese pronunciation of the character in isolation, and translations are glosses for that pronunciation):
體 → 体 (karada "body")
舊 → 旧 (kyū "old times")
當 → 当 (tō- "this-")
與 → 与 (ata-eru "give something")
變 → 変 (hen "strange")
靜 → 静 (sizuka "peaceful")
爲 → 為 (tame "reason")
眞 → 真 (makoto "truth")
In cases like that of the first two, the entire original character has essentially been replaced by a new one, independent of the original's etymology. This type, however, is relatively rare. Another approach is to essentially replace the character with a piece of it, sometimes slightly altered, as in the third and fourth characters. Finally, probably the most common type of simplification is to change one component of the character to reduce the number of strokes and/or make it easier to write, a strategy exemplified by the fifth and sixth examples. Note that, as in the case of the sixth character, the simplification may be very subtle.
In general, old character forms are identical to their traditional Chinese counterparts, but there are some exceptions. For the seventh example character, the traditional and simplified Japanese versions coexisted as different forms of the same traditional character in China, while in Japan, what is now the new character form was at that time considered a variant and rarely used. And in a few cases, like that of the eighth character, the old character form has always been considered a rare variant in China.
Historical kana usage (歴史的仮名遣 Rekishi-teki kana-zukai)
Historical kana usage is the system of kana (i.e., phonetic character) writing used in Japan prior to the post-war reforms. More specifically, it is the version of kana orthography standardized in the Meiji Period (since before that time kana usage was not standardized). It is, broadly speaking, based on the pronunciation of Japanese in the Heian Period, the time-frame in which Early Middle Japanese (on which the grammar of classical Japanese is based) was spoken. There are several differences between historical kana usage — which is also referred to as "old kana usage" (旧仮名遣 Kyū kana-zukai) — and the modern kana orthography, called "modern kana usage" (現代仮名遣 Gendai kana-zukai) or "new kana usage" (新仮名遣 Shin kana-zukai). Some of these differences apply primarily to Sino-Japanese readings of Chinese characters, while others apply primarily to native Japanese words, and still others apply equally to both groups of words.
Broadly speaking, the differences are:
H-Row (ハ行 Ha-gyō) rule
Some examples follow (old spellings are on the left, new spellings on the right; kana in parenthesis represent the pronunciation of the preceding character):
There are some exceptions to this sound change, although they are rare. They include 母（はは） (haha "mother," expected form はわ hawa), 頬（ほほ） (hoho "cheek," expected form ほお hō), 家鴨（あひる）(ahiru "domestic duck," expected form あいる airu), and 溢（あふ）れる (afure-ru "overflow," expected form あおれる aore-ru or おうれる ōre-ru; the reasons for this different result will be explained by another rule below). Sometimes, as in the case of the first two exceptions, the sound change form exists, usually with a slightly different meaning (はわ hawa is a hyper-formal and very respectful term for mother) or is used in different contexts (ほお hō is generally used in isolation, while ほほ hoho is generally used in compounds). In other cases, as is true of the second two exceptions, the unchanged form is the only one that exists. In addition to these exceptions, some dialects may preserve these sounds as they were at any stage of the language.
W-row (ワ行 Wa-gyō) rule
This section uses Nihon-shiki romanization for ゐ, ゑ, and を.Some examples:
Native Japanese words
Sino-Japanese words
There are no known exceptions (besides the aforementioned ones regarding を wo) in standard Japanese, and no dialects preserve the distinction between /wi/ and /i/, /we/ and /e/, and/or /wo/ and /o/, but some of the Ryukyuan languages (which are also descended from Proto-Japanese) do.
D-row (ダ行 Da-gyō) rule
This section uses Nihon-shiki romanization for じ, ず, ぢ, づ.Some examples:
Native words
Sino-Japanese words
Loanwords
There are no known exceptions in standard Japanese pronunciation, although there are many dialects (such as the Tosa dialect) that preserve the distinction between historical /z/ and /d/ in speech, usually by using /ʑi/ and /zu/ for historical /z/ and /d͡ʑi/ and /d͡zu/ for historical /d/ (see Yotsugana). In writing, the distinction is preserved in single morphemes in cases where a sequence ちぢ (chidi) or つづ (tsudu) was historically produced by rendaku (such as in 縮(ちぢ)む chidim-u, "shorten," and 続(つづ)く tsuduk-u, "continue", pronounced as if ちじむ chizim-u and つずく tsuzuk-u, respectively), or in compounds where a phonemic /chi/ or /tsu/ has been voiced to /zi/ or /zu/ (such as in 身(み)近(ぢか) mi-dika "one's surroundings" and 仮(か)名(な)遣(づかい)kana-dukai "kana usage," pronounced as if みじか mi-zika and かなずかい kana-zukai, respectively). This usage is a holdover from this rule.
Y-row (ヤ行 Ya-gyō) rule
In modern Japanese, the small kana ゃ/ャ, ゅ/ュ, and ょ/ョ (ya, yu, and yo) are used to indicate palatalized consonants (拗音 Yōon) when followed by an I-coloumn (イ段 I-dan) kana of the K-, G-, N-, B-, P-, M-, or R-rows (カ～, ガ～, ナ～, バ～, パ～, マ～, or ラ行; Ka-, Ga-, Na-, Ba-, Pa-, Ma-, or Ra-gyō). For example:
When a small Y-row (ヤ行 Ya-gyō) kana follows an I-column kana of the S-, Z-, T-, D-, or H-rows (サ～, ザ～, タ～, ダ～, or ハ行; Sa-, Za-, Ta-, Da-, or Ha-gyō), the preceding consonant is changed:
These three kana cannot follow A-row (ア行 A-gyō) or W-row (ワ行 Wa-gyō) kana in this way.
In historical kana, all of these examples are written with large kana や/ヤ, ゆ/ユ, and よ/ヨ (ya, yu, and yo). So the previous examples would be written:
This is the only historical kana rule that does not reflect a historical pronunciation. It is also one of only two rules (along with the geminate rule) that create ambiguity for the reader (excluding the exceptions listed above for the H-row rule). For instance, the aforementioned word 客 (kyaku) is not differentiated in historical kana from the word 規約 (kiyaku "agreement") when written in historical kana: both are written きやく (kiyaku).
Geminate (促音 Sokuon) rule
The other use of small kana in modern Japanese is in the geminate consonant mark (促音 Sokuon), っ/ッ, which is a small version of つ/ツ (tsu). In native Japanese words, this symbol can be used before kana of the K-, S-, T-, and P-rows. For example,
Voiced geminates are generally prohibited by Japanese phonological rules, but they occur in a few loanwords (although they are sometimes pronounced by native speakers as if they were their voiceless counterparts). For example:
Kana of the N- and M-rows can also be geminate, but they are preceded by ん (n) to indicate gemination instead.
Gemination can occur in Japanese for a variety of reasons. In native words, it occurs either when a historical long vowel elides, as in the aforementioned 真っ直ぐ (massugu, originally まあすぐ maasugu), or randomly, as in the aforementioned 屹度 (kitto, originally きと kito). These examples of the geminate consonant marker, along with those found in loanwords, are written with large つ (tsu) in historical kana. Therefore,
In these cases, the historical usage is not reflecting any historical pronunciation. However, in Sino-Japanese words, geminate consonants are produced by different, more regular processes, and the historical usage for these words reflects historical pronunciations.
The most common way for geminates to be produced in Sino-Japanese words is by the elision of a vowel from the kana き, く, ち, or つ (ki, ku, chi, or tsu). For example:
In historical kana, where the geminate mark is used in the first, second, and fourth examples, a full-sized version of the original kana is used. However, in the third example, つ (tsu) is used, even though an /i/ has been elided. The reason for this is that in Early Middle Japanese, when these sounds were borrowed from Middle Chinese, the Japanese language acquired a final /t/ in the Sino-Japanese morphemes that currently end in ち (chi) or つ (tsu). Later on, these acquired two forms, one with /i/ and one with /u/ (although in syllables beginning with /ni/, one form usually begins with /zi/, as is the case with 日). So the semantic difference between Sino-Japanese syllables ending in /chi/ or /tsu/ is almost always trivial, and the historical pronunciation was identical, so they were not distinguished in writing. Therefore, the previous examples would be written:
Occasionally, gemination may also result from a loss of a vowel after ふ (fu). These cases are complicated by the H-row rule, and perhaps because of that, are also written with つ in historical kana. For example,
is written
in historical kana.
While this usage does reflect a historical pronunciation, it, like the Y-row rule, produces ambiguity. Furthermore, since these vowels are elided in some compounds but not others, this usage obscures the difference in a way that is essentially impossible to predict.
While there are a few other processes that can cause geminates in Sino-Japanese words, they all apply to N- and M-row kana, and are not written differently in historical and modern kana.
Labialized consonant (合拗音 Gōyōon) rule
Starting in Early Middle Japanese, as more and more Chinese characters were borrowed into Japanese, the language acquired consonants fronted with glides. Those fronted with the palatal glide are described in the Y-row rule, but Early Middle Japanese also introduced consonants fronted with labial glides (i.e., CwV). These were far more limited in range than their palatal counterparts, however, affecting only the K- and G- rows. instead of /a/, /u/, and /o/ for the vowels of onset, like the palatal glides, the vowels of onset for the labial glides were /a/, /i/, and /e/, and used the kana わ, ゐ, and ゑ (wa, wi, and we). Finally, while the palatal glides are written with an I-column kana, the labial glides are written with a U-column (ウ段 U-dan) kana. However, when historical kana was standardized in the Meiji Period, only the syllables with historical /wa/ were indicated. Nevertheless, some classical texts may indicate the other differences, and some resources will refer to them, so it is useful to be familiar with them. This rule applies exclusively to Sino-Japanese words. Some examples:
くわ (written kuwa, but pronounced kwa) and ぐわ (written guwa, but pronounced gwa) (indicated in standard historical kana)
くゐ (written kuwi, but pronounced kwi), ぐゐ (written guwi, but pronounced gwi), くゑ (written kuwe, but pronounced kwe), and ぐゑ (written guwe but pronounced gwe) (not indicated in standard historical kana)
Labialized consonants sometimes occur in modern loanwords, and they are generally dealt with in one of two ways. Firstly, the labialized consonant may be changed from a sequence CwV to a sequence CuwV, both in writing and in speech. For example,
In other cases, they may be indicated with a U-column kana followed by a small A-row kana, indicating a labialized consonant. For example,
However, in these cases, an alternate version with large A-row kana generally exists (as it does in this case), indicating a monophthong pronunciation, and many speakers use the monophthong pronunciation regardless of how it is written.
There are no known exceptions to this rule, but some dialects (such as the Kagoshima dialect) preserve the distinction.
Classical auxiliary verb む (mu) rule
Modern Japanese has the syllabic nasal ん (n), which can represent a variety of sounds depending on what sounds come before and after it. Syllable final nasals are believed by many scholars to have existed in Proto-Japonic, but all agree that they were lost by the time of Old Japanese. They first re-appeared in Early Middle Japanese, with the introduction of Chinese loanwords ending in -n and -m. Therefore, the majority of occurrences of ん (n) in modern Japanese occur in Sino-Japanese vocabulary. Originally, syllabic n and m were phonemically and phonologically distinct, although the distinction was never written down, and was lost by Early Modern Japanese. For example,
However, some native Japanese words also have ん (n). This happens exceedingly rarely, and usually results from sound elision. An exhaustive list of all examples of all regular use Chinese characters ( Jōyō Kanji), the 2,136 characters and readings used in government documents and taught in schools, with syllabic nasal in their native Japanese readings numbers only 13 (14 readings), or 0.61%. They are
From the elision of a vowel following /m/ or /n/
From the elision of a full mora
From the preservation of an Old Japanese pre-nasalized consonant in a modern Japanese word
From abbreviation of another pronunciation on this list
From multiple processes
From some semantic (rather than phonetic) process
Of course, there are also some words with this sound that either lack Chinese characters or were coined in the modern or Early Modern Japanese eras, when ん (n) had been fully incorporated into the language. For example,
Regardless of how it came to be, the Japanese orthography lacked the character ん/ン (n) or any equivalent. Therefore, until the spelling reforms of 1900, む/ム (mu) was generally used to represent the syllabic nasal. Sometimes, this convention may be preserved by modern writers, but standard historical kana distinguishes む (mu) from ん (n).
There is one exception. In classical Japanese, there is an auxiliary verb (助動詞 jodōshi) む (mu) which indicated the volitonal. It, too, underwent vowel elision, and came to be pronounced as /m/ and then /n/. However, the conventions of standard historical kana call for this auxiliary verb (and any word derived from it) to be written with む (mu) even though they are pronounced as ん (n).
Since む (mu) is non-existent in modern Japanese, there are no dialects that preserve the distinction expressed in this rule. However, some may preserve the distinction between final /m/ and /n/.
Miscellaneous
Two other significant differences involve the way that kana are used in general, rather than which kana are used. The first is that Chinese characters in classical texts are often fully marked with ruby text (振り仮名 Furigana), especially in old laws and other very important documents. Ruby text is still widely used in modern Japanese, but only for characters with non-standard or ambiguous pronunciations, or sometimes in materials designed for children or foreigners. The second difference is that, especially in legal documents, Katakana were often used in the way that Hiragana are used in modern Japanese, to write out adjective and verb inflections, suffixes, and particles (送り仮名 Okurigana), and for the aforementioned ruby text.
Finally, kana iteration marks were far more common in classical Japanese, and sometimes used in ways that are considered completely obsolete in modern Japanese.
For an example of a major document written in the classical style, see as an example the original text of the 1890 Meiji Constitution, which is written in classical Japanese using historical kana, old character forms, kana iteration marks, and Katakana in place of Hiragana (although it lacks universal ruby text).
Conjugation table
Classical Japanese has the following verb classes and stem forms:
Table notes
The placement of the "-" (or "～" in the Japanese text) indicates where the stem of the vowel is. In other words, for a consonant-stem verb (i.e., the quadrigrade and N- and R-irregular classes), the final vowel is not considered part of the verb's root, so it is separated. However, for vowel-stem verbs (i.e., the upper and lower monograde and bigrade and K- ad S-irregular), the final vowel is considered part of the stem, except in the forms beginning in う (u), because these reflect an ancient contraction of the final vowel of the verb with that ending. The K- and S-irregular classes are also special in this regard, because they are believed to be derived from vowel-stem verbs originally, but were subject to ancient contractions that caused them to lose their final vowel in all forms, and so their final vowels are no longer considered part of their roots (and are thus separated from them), even though they are considered vowel-stem verbs.
The よ (yo) at the end of the imperative forms of upper and lower monograde and bigrade verbs and of K- and S-irregular verbs is optional in classical Japanese, although exceedingly common.
Verb class distribution
While the many conjugation classes may seem overwhelming, most of them contain few verbs. The quadrigrade and lower bigrade classes are the primary, containing about 75% and 20% of the verbs in the language, respectively. The upper bigrade class is small (about 56 non-compound verbs), but sizable enough to make an exhaustive list difficult. The other 6 classes all together contain between 22 and 28 verbs, depending on whether basic compound verbs are included or not. An exhaustive list of these follows, which verbs in the conclusive form, as is the most-common standard. Chinese character pronunciations are indicated by hiragana in parentheses following the given character. The first spelling listed for a given verb is the most common, and those that follow are alternate spellings. Some of these spellings are generally used for slightly different connotations of the same verb, while others are simple alternatives. In later reference, only the first spelling (in pre-World War II orthography) will be used, and the transcription will be based on the historical spelling. A black cell in one (or both) of the "modern" columns indicates that the modern spelling and/or transcription is the same as the pre-World War II version.
Table notes
Note that these translations are glosses, and may not reflect certain nuances or rare alternative meanings.
In addition, the translations are for the classical meaning of the verb, which may differ from the modern meaning of the verb if it has survived into modern Japanese either slightly (e.g., 着（き）る ki-ru, which meant "to wear [in general]" in classical Japanese, but means "to wear [from the waist up]" in modern Japanese), or significantly (e.g., 居（ゐ）る wi-ru, which meant "to sit" in classical Japanese, but primarily means "to be" (for animate objects) in modern Japanese). Some may have the same meaning, but a different pronunciation (e.g., 鑑（かゞ）みる kagami-ru "to learn from," which is generally pronounced and written 鑑（かんが）みる kangami-ru in modern Japanese). Also, even for those verbs which have survived with the same meaning and form, many are archaic and rarely used in modern Japanese (e.g., 嚏（ひ）る hi-ru "to sneeze," with the same modern meaning and form, but almost never used). On the other hand, some have kept the same meaning, form, and prominence into the modern language (e.g., 見（み）る mi-ru "to see," one of the oldest surviving verbs in the language and also one of the most common, both in classical and modern texts).
在すかり (imasukar-i "to exist" (honorific form) has three pronunciation variants, each of which can use either Chinese character: 在（いま）すがり/坐（いま）すがり (imasugar-i), 在そかり/坐そかり (imasokar-i), and 在そがり/坐そがり (imasogar-i).
Finally, the "modern" transcriptions are purely orthographic. For example, the modern version conclusive form of the classical verb 來（く）(k-u "to come") is 来（く）る (k-uru), but the modern form is given in the table as 来(く）(k-u), which is the way that a modern Japanese writer would write the classical Japanese word, rather than the way they would write the modern Japanese word.
Adjectives (形容詞 Keiyoushi)
Classical Japanese has the following classes of adjectives and stem forms:
Table notes
Some of these forms are very rare and used sparingly. In particular, the ～け (-ke) / ～しけ (-shike) are used almost exclusively in an ancient construction called ク語法 (Ku-gohō "Ku-grammar") which uses the irrealis for to form nouns from verbs and adjectives; e.g., 安（やす）し (Yasu-si "Peaceful") ⇒ 安（やす）け (Yasu-ke) + ～く (-ku) = 安（やす）けく (Yasukeku "Peace of mind"). The construction ～くば (-kuba) / ～しくば (-shikuba) appears to be an irrealis form ～く (-ku) / ～しく (-shiku) + particle ～ば (-ba) (since that particle usually attaches to the irrealis form), but is actually ～く (-ku) / ～しく (-shiku) + particle は (ha; modern pronunciation wa) with a sequential voicing sound change from は (ha) to ば (ba).
The compound forms are derived from continuitive form ～く (-ku) / ～しく (-shiku) + 有り (ar-i) = ～くあり (-kuar-i) / ～しくあり (-shikuar-i), which then became ～かり (-kar-i) / ～しかり (-shikar-i) by regular sound change rules from Old Japanese. The forms then follow the R-irregular conjugation type like 有り (ar-i), but lack the conclusive form.
Similarly, the basic conjugations have no imperative form. When it is used, therefore, the ～かれ (-kar-e) / ～しかれ (-shikar-e) forms are used. It is however, relatively rare, even in classical Japanese.
Adjectival verbs (形容動詞 Keiyou doushi)
There are the following classes for adjectival verbs:
Table notes
Adjectival verbs are essentially nouns combined with a copula, either ～なり (-nar-i) or ～たり (-tar-i). Which copula is used is specific to the adjectival verb in question.
The copulas are derived from directional particles に (ni) + ～有り (-ar-i) and と (to) + ～有り (-ar-i), respectively, yielding にあり (niar-i) and とあり (toar-i), respectively, which then lead to なり (nar-i) and たり (tar-i), respectively, by regular sound change rules. They therefore follow the R-irregular conjugation like 有り (ar-i).
As with adjectives, the imperative form is rare, but is used.