Comprehensible Input: How It Actually Works

In 1977, Stephen Krashen proposed something radical: you don't learn a language by studying it. You acquire it by understanding messages in that language.

He called this the Input Hypothesis. The idea is simple. If your current level is "i", you acquire language by receiving input at "i+1" — slightly above your current ability. Not two levels above. Not five. Just one step ahead.

This sounds obvious, but it contradicts how most language courses work. Traditional methods start with rules and drills. Krashen's research suggests you should start with stories.

Think about how children learn their first language. Nobody teaches a two-year-old grammar. They just hear language around them, understand most of it, and gradually figure out the rest. Adults can do the same thing — faster, because they already know how language works in general.

The practical challenge is finding the right input. For popular languages like Spanish or French, there are graded readers and simplified news sites. For less common languages — Polish, Dutch, Georgian — good material at the right level barely exists.

This is where AI changes everything. Instead of searching for content that matches your level, you can generate it. Tell the AI what you're interested in, what level you're at, and it produces a text that's exactly right. Every time.

The result: you spend less time searching for material and more time actually reading. Which is the whole point.