Japanese

	Release ja-v3
Words	50,766
English glosses	107,222
Grammar concepts	46
Kanji	10,384
Topics	20

Where the words come from

• Words, readings and glosses — JMdict, the Electronic Dictionary Research and Development Group’s Japanese–English dictionary, through jmdict-simplified. Kanji data is EDRDG’s KANJIDIC2, by the same route. Both are used under the EDRDG licence (CC BY-SA 4.0).
• Order — TUBELEX-ja, word frequencies from the subtitles of Japanese YouTube videos (NAIST; BSD-3-Clause). About a fifth of JMdict’s common words have no count there, because the corpus splits them into smaller words; they are kept, because JMdict marks them as common, and introduced as if they ranked about 5,000th.

How Japanese appears in a reply

Every word is introduced in kana, with a latin spelling beside it by default: tatemono／たてもの (building). Once you know the word, its kanji appears with the reading — 建物（たてもの） — and after eight more sightings, on its own: 建物. The whole progression, and how to ask for a reading back, is on Kanji.

Many Japanese words share a reading — はし is 橋 bridge, 箸 chopsticks and 端 edge. Your assistant logs a word with the gloss it was shown with, which is how langweave tells them apart.

Settings

As well as the shared settings, Japanese has two of its own:

In the app	Default
Latin spelling (latin_display)	Show	Show the latin spelling beside kana. Set it to Hide once you read kana.
Sightings → script unaided (kanji_bare_after)	8	Sightings of a known word, with its reading, before the reading comes off.