Conjugation of Japanese Verbs and Adjectives

On this page, I will be working exclusively with romanized Japanese because the conjugation schemes work with separate phonemes, not whole morae (in Japanese, morae are equal to syllables except that /N/ forms a separate mora; thus each full‐size kana represents one mora, as does the sokuon), especially for the uniform regular conjugation explained here. Two romanizations of Japanese will be used: the JSL romanization for the pronunciation, because it precisely represents the Japanese phonemes, and a kind of wāpuro rōmaji, more precisely, my own mix of JSL and Nihon‐shiki, for the spelling, because it precisely represents the spelling in kana. In all transcriptions and transliterations, the capital letter V shall always represent any vowel.

Due to the same phoneme–mora distinction, some verb endings will have wi, wu, we and wo, even though these are pronounced and written as i, u, e and o respectively. In particular, verbs ending in ‑Vu shall be regarded as ending in ‑Vwu.

Traditional Conjugations

Japanese verbs are traditionally grouped into three categories (sometimes the categories are referred to by their number):

1. the consonant‐stem, or godan, conjugation group verbs,
2. the vowel‐stem, or ichidan, conjugation verbs,
3. the two highly irregular verbs, suru and kuru, and their compounds; formerly there were more highly irregular verbs, which are nowadays godan.

In reality, there are more irregular verbs which follow one of the standard conjugation schemes closely but differ in a single form and its derivatives. These are counted under the respective conjugation. In addition, Japanese adjectives act like verbs in that they are conjugated rather than declined.

Before the reforms of World War Ⅱ, the written language, bungo, highly differed from the spoken language, kougo. (Among other things, the reforms defined the written language to be the same kougo [thus this page primarily aims to explain kougo]; the term bungo, however, still refers to the written language used at that time and not to the modern written language.) The verb conjugations of bungo are different from those of kougo:

1. the (single) yodan conjugation,
2. the ichidan conjugation,
3. the nidan conjugation,
4. four highly irregular verb conjugations (su/zu, ku, shinu/inu, ari/ori/haberi/imasogari).

If you encounter one of these in the modern language, do not get frightened. There is one-to-one mapping of these to the kougo conjugations:

1. yodan becomes godan,
2. ichidan becomes, well, ichidan (except for the only ‑eru verb, ‘keru’ in the meaning ‘to kick’, ‘to refuse’, which becomes godan),
3. nidan becomes ichidan (with ‑u and ‑uru stems becoming ‑eru),
4. su becomes suru (highly irregular), ‑zu becomes ‑ziru (ichidan), ku becomes kuru (highly irregular), ‑nu becomes godan, ‑ri becomes godan ‑ru.

The three main verb categories can be distinguished using this scheme:

1. if the verb is ‘suru’ in the meaning ‘to do’ or ‘kuru’ in the meaning ‘to come’, it is highly irregular;
2. if the verb ends in ‑iru or ‑eru and is not on the list of exclusions that are godan, it is ichidan;
3. otherwise, the verb is godan.

The following table is a quick reference of all kougo verb and adjective conjugations (actually, there is only one adjective conjugation), however, officially, ichidan is further divided into ‑eru and ‑iru and then each into so‐called rows according to the preceding consonant, and the godan conjugations are merged to a single godan conjugation and then similarly divided into rows: we end up with godan, ichidan ‑eru and ichidan ‑iru each being classified into a‑row, ka‑row etc.; out of the individual exceptions, only suru and kuru are singled out, while the rest are considered to belong to their ‘normal’ rows (iku/yuku to godan’s ka‑row, aru to godan’s ra‑row, kowu/towu to godan’s wa‑row).

Conjugation Basics

Originally, I was going to describe the standard approach to conjugating verbs and adjectives (much like at Wikipedia, for example), but then I remembered about Mash Satou’s uniform regular verb conjugation, re‐read the page describing it and decided to base my description on it.

In any case, we must start with deciding on the root stem of the verb or adjective:

1. for godan verbs, drop final ‑u.
2. for ichidan verbs, drop final ‑ru.
3. for highly irregular verbs, drop final ‑uru.
4. for true adjectives, drop final ‑i.

Uniform Regular Verb Conjugation

This conjugation combines godan, yodan and ichidan (Mash Satou’s original version does not work for yodan).

The main rule for conjugating using these schemes is that if the last sound of the root stem and the first sound of the ending are both vowels or both consonants, the one in the ending is dropped.

Not all forms, however, can be expressed with uniform rules. The schemes marked with an asterisk (*) only hold for everything except godan.

As you can see, the only form really special to godan is perfective, as continuative is formed from perfective by changing ‑a to ‑e and provisional is formed from perfective by adding ‑ra. The ending of the perfective form is what makes the 6 kinds of godan different:

This looks much like the individual conjugation table above, right? Note however that when forming godan perfective/continuative/etc., the last consonant of the root stem is dropped! Thus, for example, we get kaita for kaku and itta for iku.

The simplest way to learn these is:

1. Remember that if the last consonant of the root stem is voiced, then the ending becomes voiced (otherwise it stays the standard ‑ta).
2. Learn how the last consonant of the root stem changes when it is k, g or s.
3. Learn that if the last consonant of the root stem is voiced, it changes to ‑n̄, otherwise to ‑t.
4. Learn the irregular perfect forms of iku/yuku, kowu and towu.