Kagoshima Entrance Japanese common mistakes
Kagoshima Entrance Japanese · Understand the traps and how to avoid them.
When asked what the author of a 説明的文章 is claiming, the candidate answers with the topic of the passage (e.g. 'about the environment') rather than the specific assertion the author makes about it (e.g. 'small daily choices change corporate behaviour').
Under time pressure the candidate latches onto the most repeated noun in the passage and treats repetition as the main idea; they also confuse the introductory framing sentence, which often states the topic, with the conclusion that appears after the concessive turn (〜だが・しかし).
What to do instead
Before looking at the options, write one sentence in the form 'The author argues that X is Y because Z' using the final paragraph and any 〜べきだ／〜のではないだろうか sentence. Then eliminate any option that only names the topic without predicating anything about it. Practise on 5 passages, marking with a bracket the sentence that contains the author's judgement.
The candidate picks an answer option that is factually true according to the passage but does not answer the question asked (e.g. question asks 'why did the author mention A?', option states what A is).
Options are designed so that true-but-irrelevant statements feel safe; the candidate recognises the wording from the passage and stops evaluating whether it satisfies the interrogative (理由・目的・心情).
What to do instead
Underline the question's interrogative word (なぜ・どうして・どのような点で) and rewrite the question in plain form before reading options. For each option, ask 'Does this answer なぜ? or does it just describe?'. Reject options that merely restate content.
In 文学的文章 the candidate reports what a character did or said, instead of the emotion or intention the question asks about, or attributes the narrator's view to the character.
Literary passages shift between 地の文 (narrator) and dialogue/内心; candidates do not track whose perspective a sentence belongs to, and they treat emotional cues (actions, silence, weather description) as plot information.
What to do instead
Annotate the passage: mark quotation marks and 〜と思った／〜らしい boundaries and label each segment ナレーター or 人物. For emotion questions, locate the action or physical detail immediately before/after the line and convert it into an emotion word, then check the option matches that person, not another character.
The candidate translates 古文 word-by-word using modern Japanese meanings and answers content questions from that mistranslation (e.g. reading あはれ as 'pitiful' only, or をかし as 'funny').
Modern Japanese shares the same kana, so the candidate's modern intuition fires automatically; classical 形容詞・副詞 and 係り結び are not drilled to automaticity.
What to do instead
Memorise a short list of high-frequency 古語 (あはれ・をかし・うつくし・いと・げに・なむ) with their classical senses and drill 係り結び rules (ぞ・なむ・や・か → 連体形; こそ → 已然形). Before answering, identify the 主語 of each predicate, since classical Japanese frequently omits it.
On 漢文 questions the candidate reads the 返り点 in the wrong order and reconstructs a sentence with the verb applied to the wrong object.
レ点 and 一二点 are read visually left-to-right by habit, but the reading order jumps backwards; candidates also forget that 置き字 (而・於・于) are not read.
What to do instead
Physically number the characters in reading order on the page before translating (1,2,3...) using the rule: レ点 = read the lower character first; 一・二点 = after reaching 一, jump to 二 and come back. Then read the numbered sequence aloud. Do 10 short sentences this way until numbering takes under 20 seconds.
The candidate writes the wrong kanji for a homophone in the 漢字 writing section (e.g. 関心/感心, 対象/対照, 追求/追及) even though they can read the word.
Study focuses on recognising kanji in context rather than producing them from kana; the candidate has never contrasted the homophone pairs, so both spellings 'look right'.
What to do instead
Build a personal homophone table with a short collocation for each: 感心する／関心がある、対象となる／対照的、追求する／追及する. Write each pair from kana three times and check against the table. Do the same for 音読み vs 訓読み traps in the reading section (大人しい・生意気).
The candidate answers 慣用句・四字熟語 items by guessing from the literal meaning of one component (e.g. taking 油を売る as 'to sell oil') instead of the established idiomatic meaning.
Idioms are memorised as isolated lists without context sentences, so the literal reading competes with the idiomatic one under exam pressure.
What to do instead
Learn each idiom attached to a full example sentence and its body-part/keyword category (手・足・口・目). Test yourself by covering the meaning column and producing the meaning from the sentence, not from the idiom alone.
In grammar items the candidate judges a sentence by whether it 'sounds natural' and misses that the question tests a specific 文法事項 (e.g. 敬語 direction, 助詞 choice, 活用形 after 〜ので/〜のに).
Everyday spoken Japanese tolerates forms that written standard Japanese marks wrong; the candidate has no explicit checklist of what each particle or auxiliary requires.
What to do instead
For each grammar item, name the rule being tested before choosing (e.g. '尊敬語 for the other person's action → おっしゃる, not 申す'). Keep an error log of the rule names you missed and re-test only those rules weekly.
The candidate's 作文 answers the prompt loosely, drifts into a general essay about society, and never states a position with supporting reasons within the word limit.
Candidates prepare memorised essay frames and force the prompt into them; they also spend too long on the opening and run out of time before the conclusion, so the required 意見 and 理由 are missing.
What to do instead
Practise a fixed 3-part structure: 主張 (1–2 sentences) → 理由＋具体例 (2–3 sentences) → まとめ. Spend 3 minutes planning three bullet points before writing, and time-limit practice to the real allocation so the conclusion always gets written. Check 敬体/常体 consistency and 原稿用紙 rules (段落一字下げ, 句読点 at line end) on every draft.
In 聴解 the candidate answers from a keyword heard early in the audio and stops listening, missing the correction or condition that follows (〜ではなく・〜のほうが・やっぱり).
Working memory is overloaded by unfamiliar vocabulary, so the first recognisable word becomes an anchor; the candidate also fails to note the question's 条件 (who, when, which) before the audio starts.
What to do instead
Before the audio, read the question and write the condition to listen for (person/time/place). Take shorthand notes of numbers, names and any 〜けど／〜が turn, and delay choosing until the final sentence, where the decision is usually stated. Train with 1.25× speed audio and no replay.