Katarineko · Guides

Comprehensible input: the way into Japanese that doesn't feel like grinding

Comprehensible input is Japanese you can mostly understand while still meeting new things in it, and it is how language actually grows. There are two main ways to fail at Japanese: grind textbooks until you quit, or dive into native media and drown. Comprehensible input is the third way. Building it from the film you already love is what Katarineko does, and this guide shows how to walk that path with or without us.

The trap every learner knows

Content at your level is boring. Content you love is too hard.

The material you can understand (textbook dialogues, graded readers about a student named Tanaka going to the post office) rarely survives contact with your motivation. The material you care about (the film that made you want to learn, the show everyone quotes) speaks too fast, in words nobody taught you. So learners bounce between the two: study until it gets boring, immerse until it gets crushing, repeat, quit.

Most advice tells you to endure one side of it: grind for years, or "just immerse" and let ambiguity wash over you. Both work for the rare person with unusual tolerance for pain. This guide is for everyone else.

What comprehensible input is

Comprehensible input is language you can mostly understand while still meeting new things in it. When you follow a message, your brain acquires the language it rode in on, largely as a side effect. That is the core claim of linguist Stephen Krashen's input hypothesis, and it matches what generations of successful learners report: languages are acquired from massive amounts of input you understand, not from rules you memorize.

Krashen's shorthand for the sweet spot is i+1: input at your current level, plus a manageable step beyond it. A sentence where you know every word teaches you nothing new. A sentence where you know none of them teaches you nothing at all. The step in between, one new word held in place by nine familiar ones, is where acquisition happens.

What the research supports

The internet treats i+1 as either gospel or nonsense; the truth is in between. Linguists have criticized the hypothesis as hard to test. Nobody can measure exactly where your "i" is, so "+1" resists controlled experiments. Treat it as a model, not a law.

What holds up firmly is the practice the model points at. Long, sustained reading of material you chose and can mostly follow grows vocabulary and comprehension; Day and Bamford's extensive-reading work is the classic starting point. Hu and Nation put a number on "mostly follow" for one specific case: reading bare text with no help, where comfort takes roughly 98% known words. A film is a different case. Pictures, sound, and a story you chose carry meaning alongside the words, which is why a much lower share of the dialogue is enough to start following.

Three practical rules survive the academic debate. Input drives acquisition. The input must be mostly understood to do anything. And you need a lot of it, which is why the material must be interesting: nobody sustains hundreds of hours of boredom.

Why raw native content fails beginners

Now the uncomfortable part. We analyzed every line of Spirited Away: it uses 846 different words, and following most of its dialogue takes the right 214 of them. A first-year textbook teaches a few hundred words, and not necessarily these.

Here is what that gap feels like from inside. You catch a word here, a word there, and the sentence is gone. You rewind. You peek at the subtitles. Around the third rewind it stops being a film and starts being homework, and that is where most people quit. It is not a talent problem. There are simply too many words you have not met yet.

The learner-made alternatives each solve half the problem. Easy podcasts are comprehensible but they are not the story you care about. Watching tools like the ones in your stack shine once you can mostly follow. And "finish the textbook first" postpones the good part for years.

The missing piece: input derived from your destination

The problem was never that your film is too hard. The material between you and the film doesn't exist: material written at your level, made of the film's own language. So make it exist.

A film is a closed world. It uses a finite vocabulary, and it reuses it constantly. That makes the gap between you and any specific film measurable:

These are not "learn Japanese" numbers; they are "learn this film's Japanese" numbers, and they are an order of magnitude friendlier. The words also travel: finish My Neighbor Totoro's curriculum and you already cover 33% of Kiki's Delivery Service. Every film you finish makes the next one cheaper.

How Katarineko builds the path

You choose a film you want to understand. A short calibration drawn from that film's own vocabulary estimates which of its words you already know. Then, each week, you read a story written from the film's vocabulary at roughly your level plus a step. Every phrase breaks down on hover; spaced review keeps what you met. Week by week, the film's coverage climbs toward the point where watching stops being decoding.

We do not promise fluency in two months. The promise is narrower: Japanese immersion as your hobby within two months, the point where native material you chose feels enjoyable and self-sustaining. Fluency still takes years of input after that; you will be enjoying yourself for them.

How to start this week (with or without us)

The method is bigger than any tool, ours included. If you are starting from zero, pick up the basics first: our beginner resources page lists the good free options. If you want the full self-assembled version, the Refold roadmap is the best-known map of it, and here is how we fit into that method.

The short version: pick one native thing you love and commit to it. Get input daily at whatever level you are; graded readers and learner podcasts help here. Learn your film's most frequent words first. Reread and rewatch; repetition of the same world beats novelty. And drop anything that bores you. Boredom means the hours will never accumulate.

If you would rather have the measuring and the story-writing done for you, from the film you already care about, that is what we built.

Common questions

Do I need to finish a textbook before immersing?

No, but zero foundation makes everything slower. A few hundred common words and basic grammar make input much more comprehensible. Get that from any beginner resource, then let input take over. Waiting to feel “ready” is the mistake; that feeling never arrives.

Is i+1 scientifically proven?

As a precise mechanism, no. It is hard to falsify, and linguists have said so since the 1980s. As a design principle (keep input mostly comprehensible, with a little novelty) it agrees with the coverage and extensive-reading research and with what successful learners do. Treat it as a rule of thumb.

When can I start watching without subtitles?

Earlier than you think, if you pick the target first and prepare for it. For a specific film whose high-value vocabulary you have learned, 70%+ coverage is reachable in months. For “any anime, cold”, the answer is years. Choose the specific film.

Isn't 70% too low to enjoy anything?

70% of the words gets you more than 70% of the story: you also have the pictures, the music, the faces, and a plot you chose because you love it. More importantly, 70% is where the engine starts running on its own. With most of each sentence known, a new word arrives in context you understand, so watching itself begins teaching you, and comprehension climbs from there. It is not fluency and we don't pretend it is.

Keep reading

• How to start Japanese immersionThe content ladder from learner podcasts to native films, and a schedule that survives real life.
• Learn Japanese with animeWhy it fails as a starting point, and the target-first way to make the show you love actually teach you.
• Where Katarineko fits your stackWaniKani, Anki, Bunpro, tutors: what to keep, and how the pieces combine.

Or skip the theory and look at every film in the catalogue, ranked by difficulty.

Film numbers on this page are computed from full subtitle analyses; see the methodology. Research referenced: Krashen, Principles and Practice in Second Language Acquisition (1982); Hu & Nation, "Unknown vocabulary density and reading comprehension" (2000); Day & Bamford, Extensive Reading in the Second Language Classroom (1998).