JLPT Grammar Question Types Explained: Form Selection, the Star and Text Grammar

The three formats and what each one measures

The grammar block sits inside the language knowledge and reading section from N3 upwards, and it always has the same three parts in the same order. Each measures a different skill, and the list below gives the trap each is built around.

• Form selection: a gap in one sentence, four patterns; the trap is attachment and register
• Sentence ordering: four chunks, one on a star; the trap is translating word order
• Text grammar: several gaps in one passage; the trap is reading sentence by sentence

Form selection: attachment, register and nuance

Form selection gives you a sentence with a gap and four patterns. A typical N2 item: 雨が降らない ( ) 、試合は行われます, with the choices かぎり, かぎりに, にかぎって and ばかりに. All four contain kagiri or a relative, deliberately: kagiri follows a plain verb and means as long as, kagiri ni follows a noun and means starting from, ni kagitte means precisely when, and bakari ni follows a past verb and means just because, with a bad result.

The trap is threefold. Attachment: does the pattern follow a noun, a plain form, a te-form or a stem? Register: formal, so that zaru o enai fits, or conversational, so that nakya fits? Nuance: does the sentence describe bad luck, an obligation, a limit, a contrast? Look at the word before the gap first, eliminate any pattern that cannot attach to it, then read the whole sentence for nuance.

Patterns that look alike and how the test separates them

Form selection distractors are rarely random. They share a word, a sound or a translation with the answer, and attachment or nuance separates them. The table gives twelve pairs and near-pairs from N3 to N1 with the distinction that decides the point.

The levels are indicative rather than official. Many patterns return at a higher level as a nuance test in a longer sentence.

Sentence ordering: solving the star

The star format shows a sentence with four consecutive gaps, one marked with a star, and four chunks. You answer with the chunk on the star. A typical N2 item: 駅前に ___ ___ ★ ___ そうだ, with the chunks 若い人に, できたカフェは, 新しく and 人気がある. The order is atarashiku, dekita kafe wa, wakai hito ni, ninki ga aru, so the star is wakai hito ni.

The method has three steps. First, pair the chunks that must go together: an adverb with the verb it modifies, atarashiku dekita; a noun with its particle and the verb that governs it, wakai hito ni ninki ga aru. Second, build from the end backwards: sō da needs a plain form before it, so ninki ga aru goes last. Third, write the full order in the margin before choosing the star; picking it by eye is how neighbouring chunks get reversed.

Text grammar: following the writer's argument

Text grammar gives you a passage of several paragraphs, often a personal essay or a short opinion piece, with four or five gaps. More than one choice is usually grammatical inside its own sentence; the surrounding sentences decide, by whether the writer is adding, reversing or concluding, and by who the subject is.

A typical gap: つまり、問題は時間ではなく、( ) のだ, where the previous paragraph has complained about a lack of funding. The choices might be お金な, お金だ, お金の and お金で, and the answer is お金な because no da after a noun needs na. Another gap tests the connective itself: after a sentence of expectation, ( ) 、結果は予想と全く違っていた needs tokoro ga, not sono tame or sara ni. Read the whole passage once before touching a gap, then fill each by asking what the writer is doing at that point.

Connectives that decide text grammar gaps

Most text grammar gaps that are not verb forms are connectives, and the set is small. The table gives the twelve that carry most passages, with the logical job each one does. Learn them as jobs, not translations: tokoro ga and shikashi both mean however, but tokoro ga insists the outcome was unexpected.

Naze nara is the one that most often goes wrong, because it needs kara da or kara de aru at the end of its sentence; an item that offers it for a sentence without that ending is testing exactly this.

A time budget for each level

The grammar block shares its time with reading, and reading is where marks are slowest to earn, so the grammar budget should be tight. The figures below are working targets rather than official timings, which are published by the JLPT and can change.

• N3: form selection under 45 seconds per item, star items about 90 seconds, text grammar about 5 minutes
• N2: form selection under 40 seconds per item, star items about 90 seconds, text grammar about 5 minutes
• N1: form selection under 35 seconds per item, star items about 75 seconds, text grammar about 6 minutes
• All levels: mark uncertain items, move on, and return only after the reading passages are done

A form selection item you cannot solve in forty seconds will rarely be solved in two minutes, whereas two extra minutes on a reading passage often turns a guess into a correct answer.

Mistakes that cost grammar points

Four mistakes account for most lost points across the three formats. Choosing a pattern by meaning without checking attachment, so that zaru o enai is picked after a masu stem. Building the star sentence in English order, so that the modifier lands after the noun. Filling a text grammar gap from its own sentence alone, so that the connective contradicts the paragraph. And mixing registers, so that a formal essay gains a conversational nakya or a business dialogue loses its keigo.

The four sentences below each use a pattern that is frequently misattached or misplaced, correctly. Say them, then rewrite each with a different subject or tense and check that the join still holds.