Japanese Verb Groups: 一段, 五段, and Irregular

Japanese verb groups (一段 ichidan, 五段 godan, and irregular) are the three productive verb classes of modern standard Japanese. Every conjugation rule a learner meets branches on which class a verb belongs to.^{1 2} Classifying a verb correctly once lets the learner derive every later form from a fixed pattern. Misclassifying it carries the error into every downstream form.¹

Overview

Why "verb group" is the first thing you learn about a verb

The ます-form, ない-form, te-form, た-form, conditional, potential, passive, causative, volitional, and imperative all come from the verb's class. There is no morphological shortcut that bypasses the classification step.^{1 2}

Learner dictionaries include the verb-class label on every entry (Group 1 / Group 2 / Group 3, or v1 / v5 / vs / vk in JMdict-style codes) because no rule can be applied without it.¹ Both Genki and Minna no Nihongo introduce the labels in their first conjugation chapter, before teaching any conjugation paradigm.^{2 3}

Class is grammatical, not register

The three classes carry no politeness or formality level. There is no "more formal" class. All three appear at every register and in every genre in the BCCWJ (Balanced Corpus of Contemporary Written Japanese).⁴

The three classes at a glance

Modern standard Japanese has exactly three productive verb classes in the count used by learner-facing references.^{1 2 3 5 6}

Class	Stem shape	Dictionary form ends in	One example
五段 (godan, Group 1, u-verbs)	consonant-final stem	う・く・ぐ・す・つ・ぬ・ぶ・む・る	書く (write)7 8
一段 (ichidan, Group 2, ru-verbs)	vowel-final stem (い or え)	る only	食べる (eat)7 8
不規則 (irregular, Group 3)	stem shifts across forms	る (only two members)	する (do), 来る (come)9 10

The names "五段" and "一段" describe what the stem vowel does across forms. A 五段 verb cycles through all five vowel grades (あ・い・う・え・お) as it conjugates. A 一段 verb keeps one vowel for every form.^{11 12} The irregular class has just two members, and their stems shift across forms in ways neither productive class does.^{1 13}

田中たなかさんは毎朝まいあさコーヒーを飲のみます。⁸
"Tanaka-san drinks coffee every morning."

私わたしは朝あさご飯ごはんを食たべます。²
"I eat breakfast."

毎日まいにち日本語にほんごを勉強べんきょうします。¹⁰
"I study Japanese every day."

Naming conventions across textbooks

Modern Japanese verb classes use at least four parallel naming systems. All refer to the same underlying classification. No system is more "correct" than the others; each reflects a different tradition.^{1 14 11}

• Modern Japanese-language teaching (Ⅰ / Ⅱ / Ⅲ): Ⅰグループ = 五段; Ⅱグループ = 一段 (covering both 上一段 and 下一段); Ⅲグループ = irregular (する and 来る).^{3 14}
• Japanese school grammar (国語文法 / 学校文法): 五段, 上一段, 下一段, カ行変格 (kuru), サ行変格 (suru). Five classes, not three.¹⁴
• Western learner pedagogy (Genki and similar): Group 1 = u-verbs = 五段; Group 2 = ru-verbs = 一段; Group 3 = irregular.²
• Descriptive linguistics: consonant-stem verbs = 五段 (also "quinquegrade"); vowel-stem verbs = 一段 (also "unigrade"); irregular verbs.^{15 16 11 12}

The only structural mismatch for a textbook switcher is that school grammar's 上一段 and 下一段 collapse into a single class (一段 / Group 2 / ru-verb) in every modern teaching framework.^{14 11} The five-class school-grammar split survives because it tracks the classical Japanese conjugation system, not because modern Japanese needs the distinction.¹⁷

A reconciliation table that lines up every label on the same six verbs appears later in this article, in the section on how textbooks split the system differently.

The dictionary-form diagnostic

Step 1: Does the verb end in る?

If the dictionary form does not end in る, the verb is 五段 (Group 1). That is enough to classify it.^{1 2 7}

The dictionary form of any modern Japanese verb ends in one of nine kana, all from the う-row: う・く・ぐ・す・つ・ぬ・ぶ・む・る.^{5 7 18} Eight of those nine endings are clearly 五段. Only the ninth (る) is ambiguous between 五段 and 一段, which is what Steps 2 and 3 resolve.^{5 7}

Roughly 95% of new verbs a beginner encounters can be classified by ending alone, with no further analysis.⁷

毎日まいにち学校がっこうで日本語にほんごを書かきます。²
"I write Japanese at school every day."

友達ともだちと公園こうえんで話はなします。⁸
"I talk with my friend at the park."

図書館としょかんで本ほんを読よみます。²
"I read books at the library."

Step 2: If it ends in る, is the syllable before る in the い-row or え-row?

For verbs ending in る, the vowel directly before る usually decides the class. If that vowel is /i/ or /e/, the verb is probably 一段 (Group 2). If the vowel is /a/, /u/, or /o/, the verb is always 五段 (Group 1).⁷

The 一段 case is "probably" rather than "definitely" because a small set of common 五段 verbs end in -iru or -eru while still belonging to 五段. The next section covers those exceptions.^{7 19} The 五段 case has no exception: no -aru, -uru, or -oru verb in modern Japanese is 一段.^{11 7}

私わたしは毎晩まいばん九時くじに寝ねます。²
"I go to bed at 9 PM every night."

弟おとうとは六時ろくじに起おきます。⁸
"My younger brother wakes up at 6."

この道みちは駅えきまで分わかります。⁸
"I know this road as far as the station."

The 寝る example shows /e/ before る (一段). 起きる shows /i/ before る (一段). 分かる shows /a/ before る, which is a guaranteed 五段 signal.

Step 3: The two irregulars

する (to do) and 来る (kuru, to come) are the only two members of the irregular class in modern Japanese. Both must be memorised separately because their stems shift across forms in ways no other verb's stem does.^{1 13 9 10}

Compounds inherit the irregularity. Any verb of the form noun + する conjugates exactly like する (勉強する, 結婚する, 旅行する, 電話する); any verb of the form V-te + 来る conjugates exactly like 来る (持ってくる, 出てくる, 帰ってくる).^{13 10} Noun + する is also a live word-formation pattern for foreign loanwords (アップロードする, クリックする). So the irregular set is effectively unbounded in the lexicon, but still has only two morphological rules.^{13 10}

妹いもうとは毎日まいにち宿題しゅくだいをします。¹⁰
"My younger sister does her homework every day."

友達ともだちが明日あしたうちに来きます。⁹
"My friend is coming to my house tomorrow."

The 五段 / 一段 / 不規則 visual map

The three steps above combine into a single decision flow. If you memorise the flow, you can classify almost any verb from the dictionary form alone.^{7 19}

When uncertain at the final step, the ない-form diagnostic later in the article is the tie-breaker.

The る-verb trap: godan verbs that look like ichidan

Why the trap exists

A small but high-frequency set of 五段 verbs end in -iru or -eru in their dictionary form. By the "vowel before る" heuristic alone, they look like 一段 verbs. But they carry a consonant-final stem (kaer-, kir-, hashir-) and conjugate as 五段.^{11 12}

In these verbs, the r in the dictionary-form ending is part of the stem. It is not a marker of the ichidan -eru / -iru ending. The dictionary form is built by appending the う-row vowel u to a consonant-final stem (kaer + u = kaeru). The result looks like a vowel-stem verb in writing, but it is distinct in every conjugated form.^{11 12 15}

The consonant-stem class (Yodan in classical Japanese, then Godan) was already large in Old Japanese. The surface collision with 一段 verbs is a side effect of sound change between vowels in Late Middle Japanese. It is not a sign that the verbs themselves shifted class.¹⁷

The common-exceptions list

Every N5 learner is expected to memorise the high-frequency members of this set. Each verb below is paired with its English gloss and its ない-form. The ない-form demonstrates the godan diagnostic: ない attaches to the /a/-grade stem (kaer-anai, kir-anai), which a 一段 verb never does.^{7 19}

Verb	Reading	Gloss	ない-form
帰る	kaeru	to return home	帰らない7
切る	kiru	to cut	切らない7 11
走る	hashiru	to run	走らない7 19
入る	hairu	to enter	入らない7 19
知る	shiru	to know	知らない7 19
要る	iru	to need	要らない7 19
限る	kagiru	to limit	限らない19
蹴る	keru	to kick	蹴らない19
喋る	shaberu	to chat	喋らない7 19
滑る	suberu	to slide	滑らない19

Less common but still in N4 range: 焦る (aseru, "to be in a hurry"), 握る (nigiru, "to grip"), 練る (neru, "to knead"), 散る (chiru, "to scatter"), 減る (heru, "to decrease"). All conjugate as 五段.^{7 19}

父ちちは毎日まいにち七時しちじに家いえに帰かえります。²
"My father comes home at 7 every day."

紙かみをはさみで切きります。⁸
"I cut the paper with scissors."

その人ひとを知しりません。¹
"I don't know that person."

Minimal pairs to watch

Three high-frequency minimal pairs share the same kana root but split between 五段 and 一段. They therefore conjugate differently. The kanji spelling disambiguates them in writing, making it the most reliable cue for drilling the diagnostic.^{1 11}

• 帰る (kaeru, 五段, "to return home") vs 変える / 換える (kaeru, 一段, "to change / exchange").¹
• 切る (kiru, 五段, "to cut") vs 着る (kiru, 一段, "to wear (clothing on the torso)").¹¹
• 入る (hairu, 五段, "to enter") vs 入れる (ireru, 一段, "to put in"). These are different verbs with related meanings; the consonant-stem and vowel-stem split makes them conjugate as different classes.^{1 7}

The 切る vs 着る split shows the rule

Both surface as kiru in romaji and end in -iru in kana. The stem of 切る is kir- (consonant-final, godan), so the negative is 切らない. The stem of 着る is ki- (vowel-final, ichidan), so the negative is 着ない.¹¹

私わたしは十時じゅうじにうちに帰かえります。²
"I'll go home at 10."

仕事しごとを変かえました。¹
"I changed jobs."

寒さむいのでセーターを着きます。⁸
"It's cold, so I'll wear a sweater."

The ない-form diagnostic when the dictionary form is ambiguous

When the dictionary form ends in -iru or -eru and the verb is not in a memorised exception list, the ない-form is the tie-breaker. Japanese school grammar uses a simple diagnostic, taught to native speakers from middle school onward: attach ない and look at the vowel directly before ない.^{14 20}

• An /a/-row vowel before ない (-anai) signals 五段. Examples: 書か-ない, 話さ-ない, 飲ま-ない, 帰ら-ない, 切ら-ない.^{14 20}
• An /i/-row vowel before ない (-inai) signals 上一段, a subtype of modern 一段. Examples: 見-ない, 起き-ない.¹⁴
• An /e/-row vowel before ない (-enai) signals 下一段, the other subtype of modern 一段. Examples: 食べ-ない, 寝-ない, 開け-ない.¹⁴

This is the diagnostic Japanese school students learn for distinguishing the five school-grammar classes. In the three-class system used for foreign learners, the /a/-vowel result still picks out 五段, and the /i/ or /e/ result still picks out 一段, regardless of subtype.^{14 20 11}

Here is the diagnostic in action: 食べる → 食べない (/e/ before ない) → 一段. 切る → 切らない (/a/ before ない) → 五段. Once the learner can produce or guess the ない-form, every ambiguous dictionary form can be resolved.^{14 7 19}

朝あさご飯ごはんを食たべない人ひともいます。²
"Some people don't eat breakfast."

その靴くつは履はきません。切きらないでください。⁸
"I'm not wearing those shoes. Please don't cut them."

The two irregular verbs: する and 来る

する: the productive verbaliser

する ("to do, to make X happen") is the most productive verb in modern Japanese. Beyond its independent use, it forms verbal compounds by attaching directly to hundreds of nouns. These nouns come heavily from both Sino-Japanese vocabulary and foreign loanwords.^{1 10 13}

The conjugation pattern shifts the stem across three vowels. It does not cycle like a 五段 verb or stay fixed like a 一段 verb. The school-grammar paradigm gives the slots as:^{14 20}

• し- + ます / ない / ましょう / た: します, しない, しましょう, した.
• す- + る: dictionary form する.
• す- + れば: provisional すれば.
• さ- + せる / れる: causative させる, passive される.
• しろ / せよ: imperative, two competing forms inherited from different historical strata.¹⁴

Noun + する forms a single verb that conjugates as a unit on the する portion. Sino-Japanese examples: 勉強する, 結婚する, 旅行する, 電話する. Loanword examples: アップロードする, クリックする, リラックスする, キャンセルする.^{10 13}

毎日まいにち日本語にほんごの勉強べんきょうをします。³
"I study Japanese every day."

兄あには来年らいねん結婚けっこんします。¹⁰
"My older brother is getting married next year."

今日きょうは何なにもしませんでした。¹⁰
"I didn't do anything today."

Beyond the basic paradigm

A handful of sinified compounds (愛する, 訳する) inflect partly like 五段 verbs. They produce forms such as 愛します and 愛さない alongside the regular する-class slots. The full する paradigm and its edge cases belong in a dedicated treatment.^{14 13}

来る: the reading-shifting verb

来る (dictionary form kuru) is irregular because its single kanji 来 takes three different kana readings depending on the conjugation slot, while the kanji itself stays put. The shift is invisible in writing because the kanji absorbs the change. It is obvious in speech because every form sounds different.^{1 13 9}

Reading	Forms it appears in
く-	dictionary 来る (くる), provisional 来れば (くれば)14 13 9
き-	polite 来ます (きます), past 来た (きた), te-form 来て (きて)14 13 9
こ-	negative 来ない (こない), volitional 来よう (こよう), causative 来させる (こさせる), passive 来られる (こられる), imperative 来い (こい)14 13 9

Writing 来 keeps a learner safe. Reading aloud forces the learner to know which slot they are in. This is the inverse of most 一段 and 五段 irregularities, which are visible in writing but cause no reading surprises.^{13 9}

The kanji 来 derives from the same character used in classical Japanese for this verb. The modern kana-reading variation reflects the historical ka-gyō henkaku (カ行変格, "ka-row irregular") conjugation class. It had exactly one member in classical Japanese and still has exactly one member in modern Japanese.^{17 11 13}

バスがもうすぐ来きます。⁹
"The bus is coming soon."

田中たなかさんは今日きょう来こない。⁸
"Tanaka-san isn't coming today."

友達ともだちが昨日きのううちに来きた。⁸
"A friend came to my house yesterday."

Compound irregulars by adoption

Compound verbs containing する or 来る inherit the corresponding irregular conjugation on the irregular portion. The rest of the compound is morphologically inert, meaning it does not change form.^{13 10}

For する-compounds (noun + する), the noun is frozen. The entire compound is treated as an irregular verb conjugating on する. Sino-Japanese examples: 勉強する, 結婚する, 電話する, 旅行する, 練習する. Loanword examples: コピーする, クリックする, アップロードする.^{10 13}

For 来る-compounds (V-te + 来る), the te-form of another verb attaches to 来る, which then carries the irregular conjugation. Examples: 持ってくる, 出てくる, 連れてくる, 帰ってくる, やってくる.¹³

Auxiliary -てくる is a separate topic

The auxiliary use of -てくる (directional and aspectual, "coming toward the speaker" or "starting to") is a productive grammar pattern with its own semantics. It is separate from the inheritance rule above. The conjugation inheritance is regular and predictable in either reading.¹³

How textbooks split this differently

Western-textbook framing (Genki, Tobira, Minna no Nihongo)

The main English-language textbooks for adult learners (Genki, Tobira, Japanese for Busy People) and the major Japanese-published learner textbook used internationally (Minna no Nihongo) all use a three-class system. Labels differ, but the underlying classification is identical.^{2 3}

Genki (3rd edition) uses "u-verbs / ru-verbs / irregular verbs", introduced in Lesson 3 with "Group 1 / Group 2 / Group 3" as synonyms.² Minna no Nihongo Shokyū I (2nd edition) uses "Ⅰグループ / Ⅱグループ / Ⅲグループ", introduced in Lesson 14 alongside the te-form, with no English-style "u-verb / ru-verb" nicknames in the main text.³ Tobira and Japanese for Busy People both use a three-class system inherited from Genki or Minna. They switch freely between the romaji labels and the group numbers depending on the section.¹

Japan-published school grammar (国語文法 / 学校文法)

Japanese school grammar, taught from elementary school through high school as part of 国語 (Japanese language arts), uses a five-class system: 五段, 上一段, 下一段, カ行変格, サ行変格.^{14 20}

• 五段活用 cycles the vowel across all five う-row grades (書く → 書か / 書き / 書く / 書け / 書こ). Examples: 書く, 飲む, 帰る.¹⁴
• 上一段活用 keeps the stem vowel /i/ across all forms (起き-る, 見-る, 起き-ない, 見-ない). Examples: 見る, 起きる, 着る.¹⁴
• 下一段活用 keeps the stem vowel /e/ across all forms (食べ-る, 寝-る, 食べ-ない). Examples: 食べる, 寝る, 受ける.¹⁴
• カ行変格活用 has one member: 来る.¹⁴
• サ行変格活用 has one member: する.¹⁴

The 上一段 and 下一段 classes use exactly the same conjugation paradigm shape: one vowel, with no shift across forms. They differ only in which vowel sits in the stem. Foreign-learner pedagogy collapses them into a single 一段 class because no learner-facing rule branches on the い versus え distinction.^{14 11}

Linguistic framing (typology)

Descriptive linguistics analyses 一段 as vowel-stem verbs and 五段 as consonant-stem verbs. The invariant stem of an ichidan verb ends in a vowel (tab-e, m-i). The invariant stem of a godan verb ends in a consonant (kak-, yom-, kaer-). The dictionary form joins the stem to the う-row vowel u. That is what makes a consonant-stem verb end in -u in writing and a vowel-stem verb end in -ru.^{15 16 11 12}

The classical Yodan class corresponds to modern Godan. The modern volitional ending -ō (書こう, 読もう) is the post-classical innovation that promoted the system from four vowel grades (a, i, u, e) to five (a, i, u, e, o). Before that change, the volitional auxiliary -mu attached to the mizenkei (未然形, the irrealis or "not-yet" stem) /a/-grade (書か-む kakamu). Monophthongisation, the change of a vowel sequence into one long vowel, /au/ → /ō/ in Late Middle Japanese, added the fifth grade.^{17 11}

Older academic literature also uses the terms "quinquegrade" for 五段 and "unigrade" for 一段. These appear in works such as Bloch and Martin but not in modern teaching. They are useful as historical synonyms, not as required vocabulary.^{11 12}

The reconciliation table

The same six verbs map cleanly across all four naming systems.^{1 2 3 14 11}

Verb	Modern (this article)	Genki / Western	School grammar (学校文法)	Linguistic
食べる	一段 (Group 2)	ru-verb / Group 2	下一段活用	vowel-stem (unigrade)
見る	一段 (Group 2)	ru-verb / Group 2	上一段活用	vowel-stem (unigrade)
書く	五段 (Group 1)	u-verb / Group 1	五段活用	consonant-stem (quinquegrade)
話す	五段 (Group 1)	u-verb / Group 1	五段活用	consonant-stem (quinquegrade)
する	不規則 (Group 3)	irregular / Group 3	サ行変格活用	irregular (Sa-irregular)
来る	不規則 (Group 3)	irregular / Group 3	カ行変格活用	irregular (Ka-irregular)

Reading down the table, the only structural difference is that school grammar splits the two ichidan rows (食べる as 下一段, 見る as 上一段) and the two irregular rows (する as サ変, 来る as カ変). All other distinctions are relabelling.^{14 11}

Good to know

Why classical Japanese had 四段 not 五段

In classical Japanese (roughly the 8th to 16th centuries), the verbs now called 五段 inflected across four vowel rows (あ / い / う / え), not five. The mizenkei carried the volitional and negative auxiliaries on the /a/-grade (書か-む, 書か-ず). The う-row form 書こ-う did not exist because the volitional auxiliary -mu attached as -mu, not as -ō.^{17 11}

The sound shift /au/ → /ō/ during Late Middle Japanese turned 書か-む kakamu into 書か-う kakau and finally into 書こ-う kakō, introducing the fifth vowel grade and promoting the class from 四段 (yodan) to 五段 (godan).^{17 11} Any classical-Japanese reference (a 古文 textbook, a classical dictionary) uses 四段 where a modern reference uses 五段. The classification is the same. Only the count of inflectional vowel grades has shifted.^{17 11}

The "just memorise it" shortcut and why it backfires at N4

A common beginner shortcut is to memorise the polite (ます-form) and dictionary forms of each new verb as two unrelated lexical items, without classifying the verb. This works at N5, where only the ます-form is actively required.^{1 2}

The shortcut fails at N4. The te-form, ない-form, conditional, and passive all force the learner to derive forms from class membership. A learner who has not classified 帰る cannot predict that the te-form is 帰って (with the small っ promotion characteristic of 五段 -ru verbs). They may instead produce the wrong 帰てて by misclassifying it as 一段.¹

The recommended habit is to tag every new verb with its class on first encounter, whether in flashcards, in a notebook, or in a digital SRS. The investment is one second per verb at N5. The alternative is re-deriving the classification under time pressure during every N4 conjugation exercise.²

A mnemonic for the nine 五段 dictionary endings

The dictionary form of every modern Japanese verb ends in one of nine kana: う・く・ぐ・す・つ・ぬ・ぶ・む・る. Eight of these (all except る) belong unambiguously to 五段.^{5 7 18}

A mnemonic that walks the row in dictionary order is 会う・書く・泳ぐ・話す・待つ・死ぬ・遊ぶ・飲む・帰る (au, kaku, oyogu, hanasu, matsu, shinu, asobu, nomu, kaeru). It gives one canonical 五段 verb per ending, with 帰る showing the -る ending as the only ambiguous case.^{5 7} The mnemonic also works as a verb-frequency drill: each of the nine verbs ranks high in the BCCWJ corpus, so memorising the row teaches both the ending inventory and useful starter vocabulary.⁴

Are there really only two irregulars?

In modern standard Japanese, yes. Only する and 来る are irregular under the criterion used here: no other modern verb shifts stems across forms.^{1 13}

Classical Japanese had more irregular classes. ナ行変格活用 (na-gyō henkaku, "na-row irregular") covered 死ぬ and 往ぬ. 死ぬ is now a regular 五段 verb, and 往ぬ is archaic. ラ行変格活用 (ra-gyō henkaku, "ra-row irregular") covered あり, をり, 侍り, いまそかり, all of which are obsolete or have been reanalysed. The classical inventory of five irregular classes has shrunk to two by attrition.^{13 17}

One edge case is worth flagging: the verb 行く (iku, "to go") classifies as 五段 but has a one-form irregularity in the te-form and past tense. 行く becomes 行って and 行った, rather than the regular -く euphonic forms 行いて and 行いた that the rule would produce. Some references describe this as an "irregularly conjugating godan verb" rather than a fourth irregular class. Modern teaching treats 行く as 五段 with a memorised te-form, not as a new irregular.^{13 11}

See also

• Japanese Verb Classes by Aspect: Stative, Continuous, Punctual, Fourth-Class (Kindaichi 1950)
• The Japanese Verb Stem (連用形): The Masu-Stem and Its Uses
• Pitch Accent for Japanese Verbs and Adjectives: The Binary Class Rule and Conjugation Shifts
• Okurigana: When Kanji Bleeds Into Hiragana
• How Japanese Grammar Works: A Big-Picture Overview

References

Footnotes

1. Makino, Seiichi, and Michio Tsutsui. A Dictionary of Basic Japanese Grammar. The Japan Times, 1986. ISBN 978-4-7890-0454-1. "Verb (Group 1 / Group 2)" overview and conjugation entries, pp. 582–595. ↩ ↩² ↩³ ↩⁴ ↩⁵ ↩⁶ ↩⁷ ↩⁸ ↩⁹ ↩¹⁰ ↩¹¹ ↩¹² ↩¹³ ↩¹⁴ ↩¹⁵ ↩¹⁶ ↩¹⁷ ↩¹⁸ ↩¹⁹ ↩²⁰ ↩²¹

2. Banno, Eri, et al. Genki I: An Integrated Course in Elementary Japanese, 3rd edition. The Japan Times, 2020. ISBN 978-4-7890-1730-5. Lesson 3 introduces verbs under the labels "Group 1 (u-verbs)", "Group 2 (ru-verbs)", and "Group 3 (irregular verbs)". ↩ ↩² ↩³ ↩⁴ ↩⁵ ↩⁶ ↩⁷ ↩⁸ ↩⁹ ↩¹⁰ ↩¹¹ ↩¹² ↩¹³ ↩¹⁴ ↩¹⁵ ↩¹⁶ ↩¹⁷ ↩¹⁸

3. 3A Corporation. 『みんなの日本語 初級I 第2版』(Minna no Nihongo Shokyū I, 2nd ed.). スリーエーネットワーク, 2012. ISBN 978-4-88319-603-4. Verb-group labels: Ⅰグループ, Ⅱグループ, Ⅲグループ. ↩ ↩² ↩³ ↩⁴ ↩⁵ ↩⁶ ↩⁷

4. 国立国語研究所 (NINJAL). 『現代日本語書き言葉均衡コーパス』(BCCWJ, Balanced Corpus of Contemporary Written Japanese). https://clrd.ninjal.ac.jp/bccwj/ (used for frequency confirmation of high-frequency verbs cited as class exemplars: 書く, 食べる, 見る, する, 来る, 帰る, 入る). ↩ ↩²

5. Bunpro. "う-Verb (Dictionary) (JLPT N5)." https://bunpro.jp/grammar_points/%E3%81%86-Verbs (limitation: language-learning publisher; used for JLPT-level confirmation and beginner u-verb forms). ↩ ↩² ↩³ ↩⁴ ↩⁵

6. Bunpro. "る-Verb (Dictionary) (JLPT N5)." https://bunpro.jp/grammar_points/%E3%82%8B-Verbs (limitation: language-learning publisher; used for JLPT-level confirmation and beginner ru-verb forms). ↩

7. Tofugu. "Japanese Verb Conjugation Groups." https://www.tofugu.com/japanese-grammar/verb-conjugation-groups/ (limitation: language-learning publisher; used for the "vowel before る" heuristic and the standard learner-facing iru/eru exception list). ↩ ↩² ↩³ ↩⁴ ↩⁵ ↩⁶ ↩⁷ ↩⁸ ↩⁹ ↩¹⁰ ↩¹¹ ↩¹² ↩¹³ ↩¹⁴ ↩¹⁵ ↩¹⁶ ↩¹⁷ ↩¹⁸ ↩¹⁹ ↩²⁰ ↩²¹ ↩²² ↩²³

8. Practice Japanese. "Verb Groups (Godan/Ichidan/Irregular)." https://practice-japanese.com/grammar/verb-groups/ (limitation: language-learning publisher; used for representative beginner example sentences and verb-group framing). ↩ ↩² ↩³ ↩⁴ ↩⁵ ↩⁶ ↩⁷ ↩⁸ ↩⁹ ↩¹⁰ ↩¹¹

9. Bunpro. "くる (JLPT N5)." https://bunpro.jp/grammar_points/%E3%81%8F%E3%82%8B (limitation: language-learning publisher; used for 来る reading-shift confirmation at the N5 level). ↩ ↩² ↩³ ↩⁴ ↩⁵ ↩⁶ ↩⁷ ↩⁸ ↩⁹

10. Bunpro. "する (JLPT N5)." https://bunpro.jp/grammar_points/%E3%81%99%E3%82%8B (limitation: language-learning publisher; used for する compound examples at the N5 level). ↩ ↩² ↩³ ↩⁴ ↩⁵ ↩⁶ ↩⁷ ↩⁸ ↩⁹ ↩¹⁰ ↩¹¹ ↩¹²

11. Wikipedia contributors. "Japanese godan and ichidan verbs." Wikipedia, The Free Encyclopedia. https://en.wikipedia.org/wiki/Japanese_godan_and_ichidan_verbs (limitation: encyclopedic reference; used for the standard mapping between linguistic terminology (quinquegrade / unigrade / consonant-stem / vowel-stem) and the school-grammar labels). ↩ ↩² ↩³ ↩⁴ ↩⁵ ↩⁶ ↩⁷ ↩⁸ ↩⁹ ↩¹⁰ ↩¹¹ ↩¹² ↩¹³ ↩¹⁴ ↩¹⁵ ↩¹⁶ ↩¹⁷ ↩¹⁸ ↩¹⁹ ↩²⁰ ↩²¹ ↩²² ↩²³

12. Shibatani, Masayoshi. The Languages of Japan. Cambridge University Press, 1990. ISBN 978-0-521-36918-3. Chapter 7 ("The Japanese Language") covers verbal morphology and the consonant-stem / vowel-stem analysis. ↩ ↩² ↩³ ↩⁴ ↩⁵ ↩⁶

13. Wikipedia contributors. "Japanese irregular verbs." Wikipedia, The Free Encyclopedia. https://en.wikipedia.org/wiki/Japanese_irregular_verbs (limitation: encyclopedic reference; used for the inventory of historic irregulars and the 行く euphonic anomaly). ↩ ↩² ↩³ ↩⁴ ↩⁵ ↩⁶ ↩⁷ ↩⁸ ↩⁹ ↩¹⁰ ↩¹¹ ↩¹² ↩¹³ ↩¹⁴ ↩¹⁵ ↩¹⁶ ↩¹⁷ ↩¹⁸ ↩¹⁹ ↩²⁰

14. 庵功雄・松岡弘・中西久実子 ほか (Iori, Isao, et al.). 『初級を教える人のための日本語文法ハンドブック』(A Handbook of Japanese Grammar Patterns for Teachers of Beginning-Level Japanese). スリーエーネットワーク, 2000. ISBN 978-4-88319-148-0. Chapter on verb classification covers the Ⅰ / Ⅱ / Ⅲ grouping and its mapping to school-grammar 五段 / 一段 / 不規則. ↩ ↩² ↩³ ↩⁴ ↩⁵ ↩⁶ ↩⁷ ↩⁸ ↩⁹ ↩¹⁰ ↩¹¹ ↩¹² ↩¹³ ↩¹⁴ ↩¹⁵ ↩¹⁶ ↩¹⁷ ↩¹⁸ ↩¹⁹ ↩²⁰ ↩²¹ ↩²² ↩²³ ↩²⁴ ↩²⁵

15. Vance, Timothy J. An Introduction to Japanese Grammar and Communication Strategies. The Japan Times, revised edition, 1993. ISBN 978-4-7890-0698-9. Treats godan as "consonant-stem" and ichidan as "vowel-stem" throughout. ↩ ↩² ↩³

16. Jorden, Eleanor Harz, with Mari Noda. Japanese: The Spoken Language. Yale University Press, 1987. ISBN 978-0-300-03831-2. Uses "consonant verb" / "vowel verb" terminology rather than godan/ichidan. ↩ ↩²

17. Frellesvig, Bjarke. A History of the Japanese Language. Cambridge University Press, 2010. ISBN 978-0-521-65320-6. Chapters on Old Japanese verbal morphology and Late Middle Japanese sound change (Yodan → Godan). ↩ ↩² ↩³ ↩⁴ ↩⁵ ↩⁶ ↩⁷ ↩⁸

18. sljfaq.org. "What are group I and group II (also consonant and vowel) verbs?" https://www.sljfaq.org/afaq/godan.html (limitation: community FAQ; used only to confirm the linguistic framing of consonant-stem vs vowel-stem against the broader literature). ↩ ↩²

19. Sensei Japanese. "A Clear Guide to Godan, Ichidan, and Irregular Verbs." https://senseijapanese.com/japanese-verbs/a-clear-guide-to-godan-ichidan-and-irregular-verbs/ (limitation: language-learning publisher; used for the three-step diagnostic and the common-exceptions inventory). ↩ ↩² ↩³ ↩⁴ ↩⁵ ↩⁶ ↩⁷ ↩⁸ ↩⁹ ↩¹⁰ ↩¹¹ ↩¹² ↩¹³

20. 国立国語研究所 (NINJAL). 「現代日本語文法概説」 (Outline of Modern Japanese Grammar). https://www2.ninjal.ac.jp/lrc/index.php (entries on 動詞の活用, verb conjugation classes). ↩ ↩² ↩³ ↩⁴ ↩⁵