Kanagawa Entrance Japanese common mistakes
Kanagawa Entrance Japanese · Understand the traps and how to avoid them.
Writing the on-yomi when the sentence context requires the kun-yomi (e.g., reading 空 as クウ in 空が青い, or 生 in 生まれる as セイ), and marking the answer without checking whether the surrounding particles and okurigana fit the chosen reading.
Candidates memorise one reading per kanji from vocabulary lists and answer by first association rather than by testing the reading inside the sentence; okurigana (送り仮名) are treated as decoration instead of a clue to which reading is grammatical.
What to do instead
For every reading question, cover the kanji, read the whole sentence aloud with the candidate reading, and confirm the okurigana match the reading (生まれる requires ま, so セイ is impossible). Build a personal list of kanji with two common readings and one example sentence each, and drill the sentence, not the isolated character.
In kanji-writing questions, producing a character with the correct general shape but wrong component placement or a missing/extra stroke, such as writing 様 without the final stroke or confusing 博 and 薄, and submitting without comparing to the printed form.
Recognition memory is far stronger than production memory: candidates can read a kanji in a passage but have never practised writing it, and stroke-order habits formed early persist uncorrected.
What to do instead
Keep a writing notebook organised by radical (部首), writing each new kanji five times while saying the radical name aloud, and self-test by writing from the meaning only. Before finishing, check the four high-frequency confusions explicitly: 氏/紙-type radical swaps, 礻/衤, 氵/冫, and 木/扌.
Treating 語句 questions (idiom, synonym, 四字熟語, 慣用句) as free association, choosing an option because it 'sounds natural' without identifying the sentence's logical relation or the exact meaning of the target word.
Everyday spoken Japanese tolerates loose usage, so candidates rely on collocation feel; they also confuse near-synonyms that differ in register or in the direction of the meaning (e.g., 役不足 vs 力不足).
What to do instead
For each 語句 item, first write the target word's meaning in your own short phrase, then eliminate options whose meaning differs even slightly. Maintain a 20-item error notebook of 慣用句 with the body part or image that fixes the meaning (耳が痛い = criticism hurts), and review it weekly rather than rereading lists.
In grammar items, choosing a particle or conjugation by ear without checking the sentence's structure, and misidentifying 主語 and 述語 in long sentences, so that は/が, に/で, and れる/られる distinctions are decided by habit.
School grammar terms (主語・述語・修飾語, 活用形) are learned as labels to recite rather than as tools to parse a sentence, so candidates cannot apply them under time pressure.
What to do instead
Practise a fixed parsing routine: bracket each 文節, mark the 述語 first, then ask 'who/what does it?' to find the 主語, then attach modifiers. For particles, test substitution (replace で with において; if the meaning shifts to place-of-action vs means, the answer is decided). Do ten sentences a day with bracketing only, no answering.
Reading an expository passage by matching words in the question to words in the text and copying the surrounding sentence, instead of locating the paragraph that answers the question and stating the author's claim in your own words.
Keyword matching feels fast and produces plausible-looking answers; candidates also skip the 接続語 (しかし、つまり、たとえば) that signal where the argument turns or is summarised.
What to do instead
Read once for the overall claim, then for each question circle the 接続語 and the 指示語 (これ、その) in the relevant paragraph, resolve what each 指示語 points to by writing it out, and only then look at the options. Reject any option that is true in the real world but not stated by the author.
Answering 'what does the author want to say?' with a moral or personal opinion, or choosing an option that overstates the claim (always, never, completely) when the passage is qualified (多くは、必ずしも…ない).
Essay-style habits from writing tasks bleed into comprehension, and strong-sounding options feel more like a 'main idea'; candidates also miss the difference between 主張 and 具体例.
What to do instead
Label each paragraph in the margin as 主張, 例, or 対比. When two options remain, compare their strength words against the passage and eliminate the stronger one unless the text uses that strength. Practise writing the main idea in one sentence of 20–30 characters before touching the options.
In fiction, summarising the plot instead of explaining the character's feelings, and ignoring the 心情の変化 signalled by actions, weather, and silence; answers describe what happened rather than why the character acted.
Plot is easy to retell and feels like comprehension; candidates also overlook that Japanese literary passages often express emotion indirectly through 情景描写 and 比喩.
What to do instead
For each scene, draw a two-column table: できごと (what happened) and 心情 (with the line number of the evidence). Answer every feeling question with 'because [evidence line]' attached. Learn the standard signals: 目をそらす, ため息, 沈黙, and seasonal imagery usually mark a shift in feeling.
In 古文, translating word by word into modern Japanese without applying 歴史的仮名遣い rules (ゐ→い, けふ→きょう, は行→わ行) and without recognising 古語 whose modern meaning differs (あはれ、をかし、いと、うつくし).
Candidates read classical text as if it were modern Japanese with strange spelling, so they guess from surface similarity and miss the false friends that decide the meaning.
What to do instead
Before translating, rewrite the passage in 現代仮名遣い in the margin; then check every word against a short memorised list of 古語の意味 (うつくし＝かわいい, あはれ＝しみじみとした情感). For 漢文, memorise the 返り点 reading order by reciting five short passages aloud with 訓点 until the pattern is automatic.
In 漢文, reading the kanji in Chinese word order or ignoring 返り点 and 置き字, so that 不 and 未 are read as plain negatives and 而已 or 也 are pronounced as if they carried meaning.
The 訓読 system is taught as rules to memorise separately from meaning, and candidates never practise producing a full 書き下し文, so the rules never connect to comprehension.
What to do instead
For every 漢文 passage, produce a complete 書き下し文 in writing, then check it against the 返り点 one clause at a time. Keep a table of the ten most frequent 置き字 and 再読文字 (将、且、盍、未、不) with their 書き下し and meaning, and review by writing from the character only.
In the 作文 task, writing a general opinion with no personal experience, exceeding or falling short of the specified length and format (字数, 段落, 条件 such as 'experience first, then opinion'), and leaving obvious 誤字 and inconsistent 常体/敬体.
Candidates prepare memorised topic sentences and try to fit them to any prompt; time pressure means the 条件 are read once at the start and never rechecked, and style consistency is not taught as a checklist.
What to do instead
Spend three minutes before writing: underline every 条件 in the prompt, sketch a four-paragraph plan (experience → what it taught → general claim → conclusion), and assign a rough character count per paragraph. Reserve five minutes at the end to check 条件 one by one, unify 常体/敬体, and fix 送り仮名 and 誤字.