What comprehensible input actually means, and how to get it from books

Spend five minutes in any language-learning community and you'll run into the phrase "comprehensible input." CI. "Just get more input." The idea is everywhere now, but somewhere along the way it turned into a slogan instead of an actual concept.

Usually what people mean by it is something like: stop studying grammar, stop memorizing lists, and just watch things in your target language until your brain figures it out. There's some truth behind that, and honestly quite a lot of it.

But the slogan leaves out the most important part, which is that the input has to actually be comprehensible. That word is doing almost all the work.

What comprehensible input was actually supposed to mean

The idea is most associated with linguist Stephen Krashen, who proposed several hypotheses about second-language acquisition back in the 1970s and 1980s, and two of them matter here.

The first is the difference between learning and acquisition. Learning is conscious knowledge, so you study a grammar rule, you memorize that Spanish verbs change their endings, and you can explain why something is correct.

Acquisition is a different thing entirely. It's the unconscious ability to know that something sounds right, and a native speaker usually isn't thinking "this sentence uses the third-person singular present tense." They just know.

Krashen argued that these two systems are separate, and that conscious study alone can't simply transform itself into natural acquisition. That strongest version of the claim is still very much debated, and modern research generally suggests explicit instruction can help, especially when it's combined with meaningful exposure.

The second idea is the input hypothesis, and it's simple enough: you acquire language when you understand messages that are slightly beyond your current ability.

Krashen called this i+1. The "i" is your current level. The "+1" is the next small step. Not something you already understand perfectly. Not something that's completely incomprehensible either. It's that little narrow zone where you understand the message, but your brain has to do just a little bit of work.

That idea is still one of the single most useful concepts in language learning.

The problem with i+1

There's one pretty obvious problem with it though, which is that nobody knows exactly what their "i" actually is. You don't know your vocabulary size perfectly, you don't know which specific words you have and which ones are missing, and you can't just go online and download "i+1 material."

So we need some kind of practical measurement, and the best tool we have is word coverage. That's simply the percentage of words in a text that you already recognize.

Research by Paul Nation and others found that reading fiction comfortably without help usually requires very high coverage. Around 95% is often considered the minimum for reasonable comprehension, while around 98% is commonly cited as the point where reading becomes much more comfortable.

The exact number matters less than the relationship behind it. At 90% coverage unknown words are absolutely everywhere and the page becomes a constant interruption. But at 98% they're occasional enough that you're actually reading rather than decoding.

This is why beginners so often pick up a native book, get destroyed by every single sentence, and conclude that reading is impossible. The problem was never reading. Instead, the problem is that the material sits miles above their level.

Why comprehensible input became mostly a video thing

If you've heard about comprehensible input recently, you probably ran into it through videos. Channels like Dreaming Spanish helped popularize the approach by having people speak slowly while using gestures, drawings and context to carry the meaning. And it genuinely works.

I want to stop and give them some love here, because I think what they're doing is genuinely great. I've personally put hundreds of hours into their videos over the years, and they're a big part of why I take this whole approach as seriously as I do. It isn't only Spanish anymore either, since there's a French version now, built the same way by the same team. Nobody paid me to say any of that, and it's a shameless free plug. I just love what they're doing, and if you're learning Spanish or French you should go watch their stuff.

But it doesn't work because video is somehow magical. Really, it works because video hands you extra information for free. When someone points at a dog while saying the word for dog, the meaning arrives without you needing a translation at all. The image does the work.

For complete beginners that's incredibly valuable, because a person can follow a video with only a few hundred words when the visual information is filling all the gaps. A beginner book can't do any of that. A beginner text doesn't point.

Where video starts to struggle

Video is great, but it does have some real weaknesses, and the first one is density. Speech is slow compared to reading. A learner-friendly video might contain around 100 to 130 words per minute, while comfortable reading can easily run past 200. Over hundreds of hours that difference really adds up, because acquisition needs a huge amount of exposure.

The second weakness is control, because speech disappears as soon as it's said. You can rewind if you miss something, but rewinding is a conscious decision, and a lot of people just keep going and let the sentence slide past them.

Text behaves a bit differently though. The sentence just stays there in one place where you can read and examine every word. You can even slow down exactly where you need to. That ability to control the input is extremely powerful.

Reading is comprehensible input too

This is the part that tends to get overlooked, because a book is still input. The catch is that traditional reading asks for a lot of vocabulary up front, since there are no gestures or pictures there to help you.

So "just read a book in your target language" is genuinely terrible advice for a beginner, and not because reading doesn't work, but because the book isn't comprehensible yet. Technology changes that equation completely though.

Assisted reading changes everything

The famous 95 to 98% coverage numbers are all about reading without help. They answer the question "how much vocabulary do I need before I can read this completely independently?" What they don't answer is how much you need when help is instantly available. And those are two very different questions.

If you can tap a word, understand it immediately and carry on with the story, the whole experience changes. You're no longer stopping every sentence to translate, and the meaning stays completely intact.

That's much closer to what actually happens in a comprehensible-input video. The speaker supplies meaning through gestures, and an assisted reader supplies it through instant definitions and context.

The result is that material which would normally be far too difficult becomes perfectly usable. You don't have to wait until you know thousands and thousands of words before you get to read something interesting. So you can start much earlier than that.

And interest matters more than people give it credit for, because the biggest predictor of success isn't whether something is theoretically optimal. Instead, it's whether you'll actually keep doing it.

What is not comprehensible input?

A few things get mislabeled as input when they really aren't.

Input you don't understand. A podcast where you catch almost nothing isn't magically training your brain, it's mostly just noise.

Vocabulary lists. Flashcards are useful, but a word sitting on its own without context isn't the same thing as acquiring it. Instead, the word has to actually mean something to you.

Grammar drills. Exercises can help you notice patterns and produce language, which is worth something, but they aren't really input.

Constant translation. If every sentence needs to be solved as an English puzzle before you understand it, the material is almost certainly too difficult.

How to actually get comprehensible input

The practical version is pretty simple. Read and listen to things you can mostly follow, and choose things you actually care about. A thriller you can't put down will do far more for you than a respected book you're forcing yourself through.

Quantity matters too. Someone reading easier material for an hour a day will usually progress faster than someone fighting through difficult material for ten minutes.

Reread things, because the second pass through a text is dramatically easier than the first. The words are familiar, the story is already giving you context, and your brain gets another chance to connect form with meaning.

Use audio and text together when you can, since audio gives you pronunciation, rhythm and listening practice while text gives you precision and control. Each one neatly covers the other's weakness.

And stay near your level. If you're looking up a word every single sentence the material is far too hard, and if you understand everything instantly you probably need something harder.

The honest conclusion

Comprehensible input isn't magic, and it isn't a replacement for every other method either. Speaking practice matters, writing matters, grammar explanations can help and so can vocabulary study. The research doesn't support the extreme version where everything except input is completely useless.

But the central idea is very hard to argue against: people acquire language by spending a lot of time understanding meaningful language.

The mistake is thinking the answer is simply "get more input." And the real answer is to get more input you can actually understand. Then get a whole lot of it.