Classical Japanese

Overview

Japanese calls this classical form 文語 (bungo), or "literary language", as opposed to 口語 (kougo), or "spoken language". It's not nearly as simple as that, though. The differences between (modern) spoken Japanese and (modern) written Japanese are much smaller than the differences between (modern) written Japanese and (classical) written Japanese. It's a largely dead form of the language loosely comparable to Elizabethan English. Though sometimes used for ceremonial or dramatic effect, no one normally still talks or writes that way, any more than we say things like "thou hast no more brain than I have in mine elbows" when not quoting Shakespeare's Troilus and Cressida.

I'm far from an expert on classical Japanese, but here's what I know.

Example text with some analysis may be found in the Translations section of the site.

Vocabulary

Obviously, a comprehensive list of vocabulary differences would be well beyond my ability, not to mention straining the capacity of the page. However, there are a few points worth noting in particular.

The language predated widespread contact with, and thus borrowing from, English and other Western languages. Therefore, foreign loanwords, except from Chinese, were few and far between in classical Japanese, unlike the haphazard smattering of katakana words seen all over Japan today. There are any number of cases where an English-inspired term has become the preferred one, but most concepts, other than those not introduced until modern times, unsurprisingly have older native-Japanese names. Naturally, classical Japanese text would use the traditional terms. For instance, a kiss would be 口付け (kuchidzuke), or even the now-rare 接吻 (seppun), rather than the currently common キス (KISU).

As Elizabethan English had its "thee"s and "thou"s, classical Japanese had personal pronouns that have effectively gone dead in the modern language. Notable examples include the first-person pronouns 我 (ware) and 拙者 (sessha), and the second-person pronouns 汝 (nanji), 御主 (onushi), and 其方 (sonata).

Pronouns are hardly the only words that have given way to other terms. One that came up frequently was 者 (mono) to mean "person", where modern Japanese would use 人 (hito). 者 is still around, but it's less often used on its own, particularly in the spoken language. It now typically appears only as a part of a longer term, like 裏切り者 (uragiri mono) or 人気者 (ninki mono), while the kanji itself most often shows up in compounds with the pronounciation しゃ (sha), as in 医者 (isha) or 学者 (gakusha).

Like other languages, Japanese has had shifting definitions over the centuries. To give a simple example, 着物 (kimono) now refers to a specific style of clothing, but when it was originally used, the meaning of the word matched the meaning of the kanji ('wear' + 'thing'), and simply meant "clothing". Similarly, 車 (kuruma), which has come to mean "car", apparently once referred to vehicles in general.

Particles have shifted too. The classical にて (nite) was more or less equivalent to modern に (ni) or で (de). とて (tote) resembled modern といって (to itte) or と思って (to omotte). The emphatic object marker をば (woba) has gone out of use almost entirely. Another noteworthy point is that the topic particle (wa), though it existed, saw much less use than it does today. Where modern Japanese would say 私は読む (watashi wa yomu), classical would more likely use 我読む (ware yomu), leaving out the particle altogether.

Conjugations

Most of the conjugations of adjectives and verbs used in classical Japanese have changed significantly or gone out of use entirely. However, all were built on one of six basic endings. The standard ordering, with brief examples of usage and modern equivalents, follows.

未然形 (mizen-kei): Irrealis

The first base indicated that something had not (yet) happened, or was not (yet) a fact. This made it suited to forming negatives, and also to expressing hypotheticals and statements of intent.

Building the negative form added ~ず (-zu). Just ~ず on its own would end a sentence, but modifying a noun used ~ざる (-zaru) or ~ぬ (-nu) instead. I'm not sure what the distinction was between the two, except that ~ぬ didn't work with certain additional endings. Either ~ず by itself or ~ずに (-zu ni) could be used connectively, and again, I'm not sure what the distinction was.

The volitional ("it is not yet so but I intend to make it so") form added ~む (-mu), often shortened to ~ん (-n). This form still persists in the modern language in a few expressions, like 言わんばかりに (iwan bakari ni), but, like the others, has mostly died out. An equivalent negative volitional ending ~じ (-ji) was also used, in contrast with the modern language, which has no negative volitional ending in common use and instead resorts to wordier phrases when necessary. ~じ (-ji) could also function as a negative tentative, another ending that modern Japanese has largely abandoned as its own form.

Modern Japanese uses its own version of irrealis form for both negative and volitional forms, as well as few other conjugations. Although godan verbs (like 書く, kaku) use a different base form for voltional (書こ, kako) than for most other uses (書か, kaka), both are considered to be irrealis form.

~ば (-ba) resembled the modern provisional form, but when used with the irrealis was apparently restricted to cases in which the condition was not assumed to be true. In other words, it was less "if P, then Q" and more "Q would be if P were".

There were other uses, too, but that should be enough to get the general idea across.

連用形 (ren'you-kei): Conjunctive

The standard verb stem in modern Japanese retains this name. The form doesn't have any particular meaning of its own, instead serving as a base for connecting other things. It functioned more or less similarly in classical Japanese, with one of the most common functions then being the past tense.

Adding ~き (-ki) to this built the basic past tense. This became ~し (-shi) when modifying a noun, which was one of the more common uses, given that verb phrases describing nouns were—and still are in the modern language—often in past tense. However, ~き and ~し were only used when describing something the speaker directly experienced or otherwise knew to be factual. ~けり (-keri) was used instead when reporting secondhand information, changing to ~ける (-keru) when modifying a noun.

~て (-te) attached to the conjunctive form resulted in the equivalent of the familiar -te form. Modern Japanese has since added an extra degree of complexity to it by giving godan verbs two different conjunctive forms, with the one used for -te form varying depending on the final consonant of the base verb.

Again, there were other endings for this form, and even other types of past tense, but these seem to have been the most common.

終止形 (shuushi-kei): Conclusive

This was the form used to end a statement. In modern Japanese, the conclusive and attributive forms are identical for verbs and -i adjectives.

There were also a few additional endings that could be added to this form, with ~べし (-beshi) being the one most worthy of note. The pre-noun form was ~べき (-beki), changing to ~べく (-beku) when used connectively or as an adverb. Most of its forms, along with the greater part of its wide array of meanings, have been largely discarded in modern Japanese. The only common survivors I'm familiar with are べき (beki) (and rarely any other form) as a "should/must" clause ending, and べく (beku) (and rarely any other form) to connect two clauses such that the second must happen for the first to be possible.

連体形 (rentai-kei): Attributive

This was the form used when modifying a noun. In modern Japanese, the conclusive and attributive forms are identical for verbs and -i adjectives.

~ごとし (-gotoshi), the one ending of note that followed this form, indicated similarity. The pre-noun form was ~ごとき (-gotoki), and it became ~ごとく (-gotoku) when used connectively or as an adverb.

已然形 (izen-kei): Realis

Conceptually, this was the "already-so" form, where the irrealis was the "not-yet-so" form. In modern Japanese, it has been supplanted by the 仮定形 (katei-kei), or conditional form. Not many endings used the realis form, but here are two.

~ば (-ba) resembled the modern provisional form, but when used with the realis was apparently restricted to cases in which the condition was believed to be true. In other words, it was less "if P, then Q" and more "P and therefore Q".

~ども (-domo) had an "even though" sort of meaning.

命令形 (meirei-kei): Imperative

As the name suggests, the imperative makes commands. Imperative form remains in modern Japanese, though naturally with differences in the conjugation.

Adjectives

In modern Japanese, we have -i and -na adjectives, plus a few oddballs that don't fit either category. Classical Japanese had four main types and a wider variety of oddballs.

-nari Adjectives

These evolved into modern -na adjectives. The particle に (ni) and verb あり (ari), often contracted to なり (nari), followed the adjective and conjugated into the appropriate form. The most typical forms were なる (naru) before nouns and なり (nari) to end a sentence. Refer to the conjugation of あり (ari) below for more details.

Just as with the modern equivalent, following one with に (ni) instead of the usual suffix allowed it to function as an adverb.

-tari Adjectives

These have mostly died out in the modern language, but were conceptually much the same as -nari adjectives. The particle と (to) and verb あり (ari), often contracted to たり (tari), followed the adjective and conjugated into the appropriate form. The most typical forms were たる (taru) before nouns and たり (tari) to end a sentence. Refer to the conjugation of あり (ari) below for more details.

Following one with と (to) instead of the usual suffix allowed it to functon as an adverb.

-ku Adjectives

These evolved into modern -i adjectives. Conjugating one replaced the ~く (-ku) with the appropriate ending. The most common were ~き (-ki) before a noun and ~し (-shi) to end a sentence.

These adjectives could not use certain ending forms unless linked (using conjunctive form) into the being verb あり (ari), which would conjugate as usual. This final ~くあり (-ku ari) would often be contracted to ~かり (-kari), resulting in conjugations like ~から (-kara), from ~くあら (-ku ara), for the irrealis, and ~かれ (-kare), from ~くあれ (-ku are), for the imperative.

-shiku Adjectives (also -jiku)

These have joined the classical -ku adjectives as modern -i adjectives. Like -ku adjectives, they could not use certain ending forms unless linked (using conjunctive form) into the being verb あり (ari), which would conjugate as usual. This final ~くあり (-ku ari) would often be contracted to ~かり (-kari), resulting in conjugations like ~から (-kara) for the irrealis and ~かれ (-kare) for the imperative.

Note that -shiku adjectives differed from -ku adjectives only in the conclusive form, where they dropped the final ~く (-ku) entirely instead of replacing it with a second ~し (-shi).

Verbs

Modern Japanese classifies nearly every verb into one of three regular types: 五段 (godan) and two functionally identical flavors of 一段 (ichidan). Alas, classical Japanese was much more complicated, with five types of regular and four types of irregular verbs, though some categories were very small (and, to be fair, the last two types of irregular verbs still linger, largely unchanged, in the modern language).

Also note that verbs now ending in う (u) once ended in ふ (hu) instead, and conjugated accordingly. Since the shift occurred, these have been regarded as belonging to the ワ (WA) column, not the ア (A) column, which is why わ (wa) even now appears in their primary irrealis forms. Presumably ゐ (wi) and ゑ (we) were also used before they died out, unless that happened first. I'm also told that even when written with は (ha), ひ (hi), ふ (hu), and へ (he), these verb endings were still pronounced as わ (wa), い (i), う (u), and え (e), as in the modern equivalents. Presumably this relates to why the u in ou is pronounced with a distinct independent u sound when it's a verb ending, as in 追う (ou), even though the combination is pronounced with an elongated o sound in most other situations, as in 王 (ou).

四段 (yodan): Four-row

The largest category, so named due to the inflected endings covering four of the five vowel sounds (only オ (o) was left out). This group has since evolved into the modern 五段 (godan), or "five-row", category, with the changes including the addition of a second irrealis form that uses the fifth vowel.

ナ行変格 (NA-gyou henkaku): N Irregular

Irregular verbs belonging to the ナ (NA) column of the kana chart, or in other words, those ending in ぬ (nu).

In classical Japanese, this category consisted of the two verbs 死ぬ (shinu, to die), and 往ぬ or 去ぬ (inu, to leave). 死ぬ has since evolved into a regular godan verb (if the only one ending in ぬ that is at all common), while 往ぬ has gone out of use.

There was also an auxiliary verb ending ~ぬ (-nu) that used the same conjugation as these verbs.

These only differed from regular yodan verbs in using -uru rather than just -u when in attributive form and -ure rather than just -e when in realis form.

ラ行変格 (RA-gyou henkaku): R Irregular

Irregular verbs belonging to the ラ (RA) column of the kana chart, or in other words, those ending in り (ri).

In classical Japanese, this category consisted of the four verbs 有り (ari, to exist), 居り (wori, (humble) to exist), 侍り (haberi, to wait upon), and いまそかり (imasokari) or いますがり (imasugari, (respectful) to exist). The first three have become the regular godan verbs ある (aru), おる (oru), and 侍る (haberu) in the modern language, while the fourth has gone out of use.

あり also formed a variety of commonly-contracted compounds, like にあり (ni ari) and its shortened form なり (nari). These used the same conjugation.

These only differed from regular yodan verbs in ending with -i rather than -u when in conclusive form.

下一段 (shimo ichidan): Lower one-row

The name refers to the e vowel in the first kana of all inflected endings, as the エ (E) row appears below the central ウ (U) row of the traditional kana chart. The shimo ichidan category still exists in modern Japanese, with the same meaning behind the term, but due to changes in conjugation, most verbs in the modern version developed from classical shimo nidan verbs.

In classical Japanese, this category consisted of the single verb 蹴る (keru, to kick), which has since evolved into a regular godan verb.

Note that, in both classical and modern Japanese, shimo ichidan and kami ichidan verbs only differ in whether the basic form ends in -eru or -iru. The e and i remain untouched during conjugation, and the -ru portions change identically for both classes.

下二段 (shimo nidan): Lower two-row

The name refers to the e and u vowels in the first kana of the inflected endings, as the エ (E) row appears below the central ウ (U) row of the traditional kana chart.

The ヤ (YA) column has no character with the e vowel, so verbs in this group used え (e) instead. Similarly, the ワ (WA) column lacks a character with the u vowel, so these used う (u). Modern Japanese doesn't have a we character either, but ゑ (we) was alive and well back then, so that problem never came up.

Just to make things more confusing, a few of these had only a single mora in uninflected form, making it awkward to tell where the stem ended and the tail began. The worst offender was 得 (u), consisting of just a vowel sound, which meant nothing was left of the entire word in half the inflected forms. Unfortunately, that one hasn't died out completely, and continues to give people headaches as the lone shimo nidan verb in modern Japanese, with the only changes being that the conclusive form is now うる (uru) rather than just う (u) and the imperative form is えろ (ero) instead of えよ (eyo). This is why, for instance, the negative form of ありる (ariuru) is ありない (arienai). Which is appropriate in a sense since ありえない describes something that is implausible or unthinkable, or should not exist.

Many of the verbs of this category have evolved into the shimo ichidan verbs of modern Japanese. Generally speaking, taking the conjuctive form and adding ~る (-ru) results in the modern equivalent. Verbs in the ワ (WA) column need to use え (e) instead of the obsolete ゑ (we), of course, so 植う (uu, to plant) has become 植える (ueru). Less obvious is that verbs in the ハ (HA) column should be treated as if they were in the ワ (WA) column when doing this, so use え (e) here rather than へ (he), resulting in 答える (kotaeru, to answer) from 答ふ (kotau), for example. As an exception, the single-kana 経 (hu, to elapse) has become 経る (heru), presumably since the hu in this case was always pronounced as hu and not u.

Note that shimo nidan and kami nidan only differed in whether the irrealis, conjunctive, and imperative forms used e or i. Other than that, they conjugated identically.

上一段 (kami ichidan): Upper one-row

The name refers to the i vowel in the first kana of all inflected endings, as the イ (I) row appears above the central ウ (U) row of the traditional kana chart. The kami ichidan category still exists in modern Japanese, with the same meaning behind the term, and includes verbs adapted from both kami ichidan and kami nidan classical verbs.

In classical Japanese, this category consisted of about a dozen verbs, of which the following, and some of their compounds, remain in fairly common use: 着る (kiru, to wear), 見る (miru, to see), 似る (niru, to resemble), 煮る (niru, to boil), 射る (iru, to shoot), 鋳る (iru, to cast metal, etc.), 干る (hiru, to dry), and 居る (wiru, to exist). The last, normally written in kana in modern times, has become the standard living existence verb いる (iru).

The verbs such as 射る and 鋳る that used い (i) as the base were regarded as belonging to the ヤ (YA) column, but since the vowel always remained i and no yi kana existed, the y never made an appearance. All verbs in the ワ (WA) column—simply put, the ones that were ゐる (wiru)—became いる (iru) when the wi kana faded out of use. Modern Japanese classifies both groups into the ア (A-line) column.

Note that, in both classical and modern Japanese, kami ichidan and shimo ichidan verbs only differ in whether the basic form ends in -iru or -eru. The i and e remain untouched during conjugation, and the -ru portions change identically for both classes.

上二段 (kami nidan): Upper two-row

The name refers to the i and u vowels in the first kana of the inflected endings, as the イ (I) row appears above the central ウ U) row of the traditional kana chart.

The ヤ (YA) column has no character with the i vowel, so verbs in this group used い (i) instead. However, the kami nidan category had no verbs in the ワ (WA) column, so the lack of a wu didn't matter.

Most of the verbs of this category have evolved into kami ichidan verbs in modern Japanese. Generally speaking, taking the conjuctive form and adding ~る (-ru) results in the modern equivalent for these. Less obvious is that verbs in the ハ (HA) column should be treated as if they were in the ワ (WA) column when doing this, so use い (i) here rather than ひ (hi), resulting in 強いる (shiiru) from 強ふ (shiu), for example. As an exception, the single-kana 干 (hu, to dry) has become 干る (hiru), presumably since the hu in this case was always pronounced as hu and not u. A stranger exception, 恨む (uramu, to resent), instead evolved into a yodan verb and so wound up in the modern godan category.

Note that kami nidan and shimo nidan only different in whether the irrealis, conjunctive, and imperative forms used i or e. Other than that, they conjugated identically.

カ行変格 (ka-gyou henkaku): K Irregular

Irregular verbs belonging to the カ (KA) column of the kana chart. The only one in classical Japanese was the verb 来 (ku, to come). This has become 来る (kuru) in modern Japanese, and remains the only verb in the category.

In modern Japanese, the conclusive form has become 来る (くる, kuru) and the imperative form 来い (こい, koi), but the other forms haven't changed. This can particularly confuse learners of the language, who may very well be told (since it's true for the regular verbs) that conjugation only affects the final kana of a verb and that the kanji part of a verb never changes. 来る defies both expectations, with くる (kuru) somehow turning into こない (konai) and きます (kimasu), all with the same kanji read three different ways.

サ行変格 (sa-gyou henkaku): S Irregular

Irregular verbs belonging to the サ (SA) column of the kana chart, or in other words, those ending in す (su).

In classical Japanese, this category consisted of the two verbs す (su, to do), and おはす (ohasu, to be, come, or go), in addition to their compounds. おはす has since gone out of use, and す has become する (suru), now standing alone as the only verb in this category in modern Japanese. And if anything, it's even more irregular now, thanks to several of its compounded words conjugating somewhat differently from both the base verb and each other.