Middle Chinese → Japanese
This is the ladder this app is built to walk you down: a language followed through its own history, with branches taken at the point where they actually happened. Because a language is the largest thing you can learn at once, the roots pay you back several times over.
Japanese takes its writing system and most of its learned vocabulary from Chinese, in waves: Go-on from the fifth century, Kan-on from Tang Chang'an, Tō-on from the Song and Ming. The Chinese characters carry meaning straight across, and the readings are fossils of which wave a word arrived in. What does not cross is the grammar.| Stage | What it is | Status |
|---|---|---|
| Middle Chinese | the branching point: the Qieyun rhyme dictionary of 601 CE is a genuine phonetic record | attested via rhyme books |
| Go-on | the oldest reading layer, arriving with Buddhism | attested |
| Kan-on | readings from Tang-era Chang'an | attested |
| Tō-on | Song and Ming readings; why one character can have several | attested |
| Japanese today | what this app can feed you now, from news text | living |
What is built, and what is not
Built now: modern Japanese news text, scored against your own words, with the letters you cannot yet read called out. That is the end of the ladder above and it is real.
Not built: the historical rungs. Reading the Rig Veda, or the Cantar de Mio Cid, or a Tang-era text, needs those corpora prepared the way the modern one is — segmented, ranked, and measured against a reader's words. The corpora are catalogued (GRETIL for Sanskrit, PHI for Latin, Kanseki and CTEXT for Chinese) but the work is not done, so there are no pages for them. When they exist, the ladder above is the order you would climb it in.
One honest caveat about the root: a root is not a shortcut to fluency in one descendant. It is the cheapest route into a whole family, and it is the difference between learning three related languages and learning one language three times. If you only ever want Spanish, going straight at Spanish beats Latin.