← Your Japanese stack

Refold and Katarineko: the method, with the grind removed

If you found us, you've probably read the Refold roadmap. Good: it's the best-known articulation of how immersion learning works, and we agree with its core. Katarineko is an engine for the roadmap's hardest stretch, not a rival method.

Where we agree

Refold's spine is the same research-backed bet our cornerstone guide makes: acquire through massive comprehensible input, and build comprehension before output. If the roadmap is your map, keep it.

The stretch where people fall off

The method's hardest miles are the early-intermediate ones. Native content is still mostly fog, and tolerating that ambiguity takes willpower. The standard toolkit asks you to assemble your own machine: mine sentences from a show into a deck, then keep the deck fed.

Refold tells you what to do; the doing is where the attrition is. Every immersion Discord has people who agree with the method and stalled in this stretch anyway.

What Katarineko automates in it

Where the roadmap says "get comprehensible input daily, aimed at content you love", we generate that. You pick the film and calibration measures which of its words you know. Weekly stories arrive written from the film's vocabulary just above your level, with ambiguity kept at the productive dose.

The leveling and the "what do I read today" question are handled. When you reach the film, the roadmap continues as written: watch natively, and mine from material you already three-quarters understand.

One divergence: classic Refold practice leans on raw native input plus separate study, trusting you to endure the fog. We derive the input from the destination instead. Purists may prefer the unassisted route, and it also works. We exist for the people the fog was filtering out.

Katarineko is unaffiliated with Refold; statements here describe its widely known roadmap. More pairings: your Japanese stack.