“Comprehensible input” is one of the most important ideas in language learning — and one of the most loosely used. Some people use it to mean “input that's easier than native content.” Others mean “graded material designed for learners.” I use a narrower definition: strict comprehensible input.
Strict CI: understand everything except one new thing
In strict CI, the learner understands every element in a message except for a single new one. That's the “+1” in Krashen's i+1: input that is just one step beyond your current level.
The new element might be:
- A vocabulary word you haven't seen before
- A grammar pattern built from words you already know
- A new conjugation or form of a verb you've encountered in another tense
- A compound word or phrase where only one part is unfamiliar
Examples of strict vs approximate CI
Suppose you know these Spanish words: hola, yo, quiero, agua.
Strict CI: “Hola. Yo quiero leche.” — You understand the greeting, the subject, the verb, and the structure. Only leche (milk) is new.
Approximate CI: “Hola. Me gustaría un vaso de leche, por favor.” — You might guess the meaning from context, but me gustaría, vaso, and por favor are all new. That's three unknowns, not one.
Approximate CI isn't useless — context helps. But it's a different experience from strict CI, where your brain can focus entirely on acquiring that single new element without the cognitive load of decoding multiple unknowns.
Why strictness matters
When input is only mostly comprehensible, several things happen:
- You spend energy guessing instead of acquiring
- You may miss the new element entirely because you're overwhelmed
- You develop a habit of tolerating confusion rather than building stable comprehension
- Frustration rises, especially for beginners
Strict CI keeps you in a zone where learning feels possible — not easy, but comprehensible. That's the whole point.
The problem: strict CI is hard to deliver with static content
Graded video series, textbooks, and reader apps can approximate i+1 for an average learner following a fixed path. But they can't know whether you already learned leche from Anki last week, or whether you skipped episode 12 and missed a grammar pattern episode 40 assumes you know.
The result: content that was strict CI for the curriculum's assumptions becomes approximate CI for you personally. Read more about static CI limits.
What makes strict CI possible for an individual learner
To deliver strict CI to you, a system needs two things:
- A model of your knowledge — what words and patterns you've actually encountered and retained
- Dynamic generation — the ability to create the next sentence using only that knowledge, plus one new element
A flashcard deck is the closest practical approximation of that knowledge model most learners already have. That's the insight behind CI+O and flashcards-to-conversation practice.
Summary: Strict CI = understand everything except one new element. Approximate CI = mostly understandable with multiple unknowns. Strict CI is more effective for acquisition, but requires personalization that static content can't fully provide.