Learning Japanese From Subtitled Video

In most languages, reading the subtitle is trivially easier than hearing the line. In Japanese it can be harder. A subtitle line like '臨機応変に対応する' requires knowing four kanji to read at all, even if you would understand 'rinki ōhen ni taiō suru' spoken. This makes synchronized subtitles doubly useful (every line is listening and reading at once), but it also means you need instant lookups that show the reading, not just the meaning. Nativra's word click returns the reading (kana) and the meaning, and the CEFR-style level tag maps roughly to JLPT bands for Japanese.

Anime and drama are conversational and often casual or stylised; they teach plain forms, sentence-final particles, and slang, but not how to speak to a colleague. News and documentaries are formal (です/ます, honorific vocabulary). Interviews sit between. Mix them deliberately: a learner who studies only anime sounds strange in a workplace; one who studies only news cannot follow friends. Comprehensible Japanese and Nihongo con Teppei are designed for learners and are the easiest way in; from there, Japanese YouTubers with captions (many use them for accessibility) and NHK material.

• A1–A2: Comprehensible Japanese, Nihongo con Teppei (beginner series).
• A2–B1: NHK News Web Easy, Japanese Ammo with Misa, Onomappu.
• B1–B2: Japanese vloggers with captions, NHK documentaries, Terrace House (subtitle file).
• B2–C1: Business interviews, NHK regular news, Midnight Diner (subtitle file).

Two-pass method with short segments. On Japanese cards, keep three fields: the word as written (kanji), its reading, and the meaning: the exported TSV supports this via the word and lemma columns. Consider suspending cards for words whose kanji you cannot yet write and revisiting them; recognition first, production later. Pitch accent is not shown by Nativra; if it matters to you, Migaku's Japanese tooling is stronger there, as the comparison page notes.