Breaking Through the Invisible Wall: School Grammar vs. Pedagogical Grammar, a Strategy Report for 2026 JLPT Candidates

If you are preparing for the 2026 Japanese-Language Proficiency Test, you may know this plateau well.

N5 and N4 went smoothly. Minna no Nihongo or Standard Japanese carried you along, and the sentence patterns came out on demand. Then N2/N1 preparation began, and something broke. A reading passage runs three or four lines before reaching its verb. A grammar entry like 「～ざるを得ない (zaru o enai — cannot help but…)」 or 「～まじき (majiki — that ought never to be)」 appears, and everything you knew seems to stop working.

The cause of that discontinuity is a collision between two operating systems: the pedagogical grammar (日本語教育文法, nihongo kyōiku bunpō) you have been learning with, and the school grammar (学校文法, gakkō bunpō) that the upper levels of the JLPT quietly assume.

This report dissects the difference across three axes — parts of speech, conjugation, and syntax — and turns each one into something you can use on the actual test.

---

1. Where the Two Systems Come From, and What the JLPT Is Really Testing

1.1 Pedagogical grammar (Kyōiku Bunpō)

• Designed for: learners whose first language is not Japanese.
• Goal: the shortest path to production. Complex morphology is packaged into fixed formulas — sentence patterns (文型, bunkei). Call it deliberate black-boxing.
• Method: morphological analysis is suppressed in favour of syntactic scaffolding. It never explains why adding 「て」 produces a linking function; it simply drills the sound changes of the "te-form."
• Standing on the JLPT: it owns N5, N4, and most of N3 outright.

1.2 School grammar (Gakkō Bunpō)

• Designed for: native speakers. Its skeleton is the theory of the linguist Hashimoto Shinkichi (橋本進吉), organised after the war into the grammar Japanese children study in 国語 (kokugo) class.
• Goal: exhaustive analysis. Japanese schoolchildren already speak the language, so patterns are useless to them. What they need is a scalpel precise enough to dissect their own speech — and, further down the road, to read classical Japanese (古文, kobun).
• Method: a microscope. Every kana-level alternation (活用, katsuyō) and every particle and auxiliary is given an independent identity.
• Standing on the JLPT: it is the logic underneath N2/N1 reading passages, upper-level word formation, and every surviving fragment of the classical language (文語の名残, bungo no nagori).

1.3 The system conflict

The JLPT is a test for foreigners, but its N2 and N1 reading passages are drawn broadly from real Japanese editorials, criticism, essays and academic prose (some adapted for the test). That prose was written by authors for whom school grammar is an unexamined default.

Bring only the building-block method of pedagogical grammar to a sentence layering causative-passives, transitive/intransitive pairs, double negation and classical auxiliaries, and you simply do not have a tool that takes it apart.

---

2. Parts of Speech (品詞分類, Hinshi Bunrui) — Where the Systems Diverge

2.1 The whole map: school grammar's ten parts of speech

School grammar first splits every word into independent words (自立語, jiritsugo — able to form a 文節 on their own) and dependent words (付属語, fuzokugo), then subdivides into ten classes. Pedagogical grammar kept only the labels it needed: noun, verb, i-adjective, na-adjective, adverb, particle.

Watch the bolded 助動詞 (jodōshi, auxiliary verbs). Pedagogical grammar all but dissolves this category — yet in N1 reading it is precisely the class that carries negation, conjecture, passivity, causation and hearsay, which is to say the author's entire stance.

The 連体詞 (rentaishi, prenominal) class deserves a mention too. 「あらゆる (arayuru, every)」, 「いわゆる (iwayuru, so-called)」 and 「大きな (ōkina, big)」 are non-inflecting words in their own right, not adjectives. That single fact explains an asymmetry every learner trips on: 「大きい」 negates to 「大きくない」, while 「大きな」 can never do so.

2.2 The quiet war over verb classes

Pedagogical grammar sorts verbs into Groups Ⅰ, Ⅱ and Ⅲ — a scheme built to derive the masu-form and te-form fast. School grammar instead names each class after which row of the kana chart the ending moves through.

There is a beautiful minimal pair that makes the "what lives in the stem" question concrete: 「切る」 and 「着る」. Both read kiru. 「切る」 has the stem 「き」 and is ra-row godan (切らない, 切ります); 「着る」 has no separable stem and is kami-ichidan (着ない, 着ます). Two identical pronunciations, two entirely different conjugations — and no way to explain the difference without school grammar's notion of a stem.

2.3 The great misunderstanding about 「な」

• The pedagogical view

  • i-adjectives: 「高い (takai)」.
  • na-adjectives: 「静か (shizuka)」, which takes 「な」 before a noun.
  • The trap: learners come to believe 「な」 is either part of the word or an ordinary particle.

• The school-grammar view (required for N2/N1)

  • 形容詞 (keiyōshi) = i-adjectives.
  • 形容動詞 (keiyōdōshi) = na-adjectives. Here is the key: 「静か」 is only the stem (語幹). The base form — the 終止形 — is 「静かだ (shizuka da)」, and that 「だ」 is not a separate word glued on. It is the word's own inflectional ending.
  • Being an ending, it changes with the word's role in the sentence: 未然 静かだろ / 連用 静かだっ・静かで・静かに / 終止 静かだ / 連体 静かな / 仮定 静かなら. So the 「な」 that shows up before a noun is not a particle at all — it is the attributive (連体形, rentaikei) ending of a 形容動詞.

Putting it to work at N2

Why does formal prose keep saying 「静かである (shizuka de aru)」?

「である」 is a compound — the 連用形 ending 「で」 plus the auxiliary verb 「ある」 — and its job is to serve as the sentence-final form of written, academic-register Japanese. Spoken: 「この部屋は静かだ (kono heya wa shizuka da)」. In a paper: 「この部屋は静かである (kono heya wa shizuka de aru)」.

And here is a trap with a very high hit rate. 「である」 replaces the 終止形 only. It does not replace the 連体形. To modify a noun you must return to 「な」: 「静かな部屋 (shizuka na heya)」 is the only option, and 「静かである部屋」 is simply wrong. (「である」 does precede nouns directly, but only when it follows a noun itself, as in 「日本人である私 (nihonjin de aru watashi), I, who am Japanese」.)

That is why 「である」 saturates N2/N1 reading passages while appearing almost exclusively at the end of sentences.

---

3. Conjugation (活用形, Katsuyōkei) — the Real Unlock

3.1 Pedagogical grammar names forms by function (fine for N5–N4)

The verb and whatever follows it are bundled and labelled by use:

1. 辞書形 (jishokei): 食べる (taberu) — dictionary form
2. ます形 (masukei): 食べます (tabemasu) — polite
3. ない形 (naikei): 食べない (tabenai) — negative
4. て形 (tekei): 食べて (tabete) — linking
5. た形 (takei): 食べた (tabeta) — past
6. 意向形 (ikōkei): 食べよう (tabeyō) — volitional

3.2 School grammar's six forms (the N3–N1 awakening)

In school grammar the verb itself has only six forms. Each one summons a different 助動詞 (jodōshi), and it is the auxiliary — not the verb — that supplies negation, past, passive or causative meaning.

Take the godan verb 「書く (kaku, to write)」:

1. 未然形 (mizenkei — the not-yet form): 書か (kaka) / 書こ (kako)

   • negative auxiliary 「ない / ぬ」 → 書かない
   • passive auxiliary 「れる」 → 書かれる
   • causative auxiliary 「せる」 → 書かせる
   • volitional auxiliary 「う」 → 書こう (「書こ」 counts as a mizenkei variant)

2. 連用形 (ren'yōkei — the conjunctive form): 書き (kaki) / 書い (kai, sound-changed)

   • polite auxiliary 「ます」 → 書きます
   • past auxiliary 「た」 → 書いた
   • conjunctive particle 「て」 → 書いて

3. 終止形 (shūshikei — the terminal form): 書く (kaku) — ends the sentence.

4. 連体形 (rentaikei — the attributive form): 書く (kaku) — precedes a noun: 「書く人 (kaku hito), the person who writes」. Identical to the terminal form in modern Japanese; distinct in classical.

5. 仮定形 (kateikei — the hypothetical form): 書け (kake) — + 「ば」 → 書けば (if one writes)

6. 命令形 (meireikei — the imperative): 書け (kake) — 「書け！」

3.3 The big payoff: dismantling classical residue

At N1 you meet grammar no one uses in conversation but editorials use constantly. Pedagogical grammar has no answer beyond memorisation. School grammar disposes of it in one step.

A classic N1 item

彼は、真実を（ 　 ）を得なかった。 A. 語らない　B. 語りざる　C. 語らざる　D. 語るざる

• Without school grammar: what even is 「ざる」? It was never in the nai-form drills.
• With school grammar:
  • The pattern is 「～ざるを得ない (zaru o enai, cannot help but…)」.
  • 「ざる」 is the 連体形 of the classical negative auxiliary 「ず (zu)」.
  • 「ず」 — like modern 「ない」 — attaches to the 未然形.
  • The mizenkei of 「語る (kataru)」 shifts the ending to the a step: 「語ら (katara)」.
  • Answer: C. 語らざる (katarazaru). In full — 「彼は、真実を語らざるを得なかった (kare wa, shinjitsu o katarazaru o enakatta), he had no choice but to speak the truth.」

The same tool opens the rest of the N1 classical set:

• 「～んがため (n ga tame)」, in order to: mizenkei + the classical volitional 「む」 in its nasalised form 「ん」. Example: 「勝たんがため (katan ga tame), in order to win」 — 「勝つ」 → mizenkei 「勝た」.
• 「～べからず (bekarazu)」, must not: the 終止形 (連体形 for ra-row irregulars) + 「べから」, the mizenkei of the auxiliary 「べし」, + the negative 「ず」. Example: 「この場に立ち入るべからず (kono ba ni tachiiru bekarazu), no entry」.
• 「～まじき (majiki)」, that ought never to be: the 連体形 of 「まじ」, the negative counterpart of 「べし」. Example: 「あるまじき行為 (arumajiki kōi), conduct that should never occur」.

---

4. Syntax — Cutting Through the Labyrinth

In N2/N1 読解 (dokkai, reading comprehension), what decides the outcome is rarely vocabulary size. It is whether you can segment a long sentence.

4.1 The SOV building-block method, and its ceiling

• [topic] は [time] に [place] で [object] を [verb].
• 私は (Watashi wa) / 明日 (ashita) / 学校で (gakkō de) / 日本語を (nihongo o) / 勉強します (benkyō shimasu).

Extremely efficient for simple sentences. It collapses the moment you hit stacked modification and embedded clauses.

4.2 The 文節 (bunsetsu) blade

The core reading drill in Japanese middle schools is segmenting a sentence into 文節.

One 文節 = one independent word + any number of dependent words. It is the smallest natural break, phonologically and semantically. The traditional test is to try inserting 「ね (ne)」 at each candidate boundary.

Let us take the blade to a sentence of genuine N1 difficulty:

The sentence グローバル化が進展しつつある現代社会において、異文化を理解しようとする姿勢を持たない者は、真の国際人とは言えまい。 (Gurōbaruka ga shinten shitsutsu aru gendai shakai ni oite, ibunka o rikai shiyō to suru shisei o motanai mono wa, shin no kokusaijin to wa iemai.)

What goes wrong: the eye sweeps up グローバル化, 進展, 現代社会, 異文化, 理解, 国際人 and reassembles them in the reader's native word order — losing the decisive negation and the conjectural mood at the end.

Segmented

1. グローバル化が (noun + particle)
2. 進展しつつ + ある (verb ren'yōkei + conjunctive particle / the auxiliary verb 「ある」. Strictly, school grammar gives a supporting verb its own 文節, so the two together form a 連文節) — which is advancing
3. 現代社会において、(noun + compound particle) — end of the scene-setting adverbial
4. 異文化を (noun + particle)
5. 理解しようと (sa-hen mizenkei 「し」 + volitional 「よう」 + particle 「と」) — seeking to understand
6. する (sa-hen verb, modifying 「姿勢」, therefore the 連体形)
7. 姿勢を (noun + particle) — an attitude
8. 持たない (verb mizenkei + negative auxiliary) — end of a very long attributive clause
9. 者は、(noun + particle) — the actual subject finally arrives
10. 真の (noun + particle)
11. 国際人とは (noun + particle + particle)
12. 言えまい。(mizenkei of the potential verb 「言える」 + the negative-conjectural auxiliary 「まい」) — the predicate, carrying the author's judgement

The takeaway

In JLPT reading, the author's position lives in the sentence-final auxiliary. Here that is 「まい (mai)」, school grammar's negative-conjectural auxiliary (pedagogical grammar's 「～ないだろう」). Its attachment is rule-governed: 終止形 for godan verbs (行くまい), 未然形 for ichidan and the two irregulars (見まい / 来まい / すまい). 「言える」 is a shimo-ichidan potential verb, hence the mizenkei 「言え」.

Segment first, and the skeleton appears in a second: 「者は … 国際人とは言えまい」 — such a person cannot be called a true internationalist.

---

5. A Level-by-Level Plan for 2026

N5–N4: stay inside pedagogical grammar, build reflexes

If you are sitting N5 or N4 in 2026, do not touch school grammar.

• Plan: follow Minna no Nihongo or Standard Japanese exactly as written, and drill te-form, ta-form and nai-form until they are muscle memory.
• Why: the job at this stage is comprehension and simple production. "Mizenkei" and "ren'yōkei" are pure cognitive load for a beginner, and a common reason people quit.

N3: the crossover — start splitting morphemes

N3 is the watershed. Transitive/intransitive pairs and long attributive clauses begin to appear.

• Plan: keep learning patterns, but start looking behind them.
• Actions
  1. Stop treating 「飲みます」 as one unit. 「飲み (nomi)」 is the ren'yōkei; 「ます」 is an attached politeness auxiliary.
  2. Stop memorising the causative-passive (食べさせられる, tabesaserareru) as a long string. Split it: 「食べる」 mizenkei 「食べ」 + causative auxiliary 「させる」 in its mizenkei 「させ」 + passive auxiliary 「られる」 (tabe + sase + rareru). Split apart, it is a trivial puzzle.

N2–N1: switch fully to school grammar

At this level pedagogical grammar has done its job. Upper-level reading will keep leaning toward argumentation and structural complexity.

• Plan: make the six conjugation forms and the auxiliary-verb table your primary weapons in the final stretch.
• Actions
  1. Rewrite your grammar notes. When working through Shin Kanzen Master or Nihongo Sō-Matome, stop recording "attaches to the nai-form minus nai" and record "attaches to the 未然形."
  2. Segment long sentences with the 「ね」 test, in pencil or in your head. Strip the modifiers, find the subject and the sentence-final auxiliary.
  3. Apply it to vocabulary. For an unfamiliar reading, try deriving it from sound-change and ren'yōkei regularities.

---

Closing

Language is a tool for communication and, at the same time, a vessel for thought.

Pedagogical grammar built you a bridge into Japanese and got you aboard quickly — nothing does that better. School grammar is the schematic that continues into the driver's cab, showing you why this precise machine runs the way it does.

If you are sitting the JLPT in 2026, there is no reason to be intimidated by competing grammatical traditions. Use both: pedagogical patterns for the split-second recognition that 聴解 (chōkai, listening) demands, and school grammar's scalpel for the long prose of 読解 (dokkai, reading).

Once you can see the morphological meaning behind every 「て」, every 「れる」 and every 「ざる」, the invisible wall in front of you comes down. Good luck with your preparation.