Listening: Extract the Task, Not Every Word
Listening sections
N4 listening has task-based comprehension, point comprehension, summary, immediate response, and integrated tasks. The skill is not translating every word; it is extracting the information the question asks for-time, place, action, number, reason, or next step. Before the audio starts, read the question and options, underline the category (when/where/who/how much), and predict the kind of word you need.
What to train
Numbers and counters (一つ, 一枚, 三人, 何回), clock times and dates, days of the week, prices, and durations. Directions and positions (右, 左, 前, 後ろ, 隣, 間, 向かい). Request and permission patterns (～てください, ～てもいい, ～なければならない, ～ないでください). Sound changes: てしまう→ちゃう, では→じゃ, なければ→なきゃ, という→って. Fillers and repairs (ええと, あ、そうではなくて) often signal a correction-the final answer is usually after the repair.
Traps
Options may repeat words from the audio while reversing the relationship. Negation can be softened (～ないことはない) or doubled. A speaker may mention two times and reject one. Polite forms can hide the actor: ～ていただけませんか is a request, not a question about ability. In immediate-response items, listen to intonation and sentence-final particles; ～ませんか invites, while ～ません declares.
Rule of thumb
Take notes in symbols, not full sentences: time as 3:00, place as a circle, action as a verb arrow. Example: a man says 「会議は三時からですが、資料は二時半までに願いします。」 The question asks when to submit materials. The meeting starts at three, but the deadline is 2:30-answer 二時半, not 三時.