Comprehensible Input according to GPT-4

The Comprehensible Input approach, based on the theories of linguist Stephen Krashen, is indeed an effective way to learn a language for many people. The main idea behind this approach is that learners acquire a new language by being exposed to it in context, where they can understand the general meaning even if they don’t know every word or grammatical structure. By focusing on messages they can comprehend, learners can gradually absorb the language’s vocabulary, grammar, and other features naturally over time.
Some advantages of the Comprehensible Input approach include:

  1. Contextual learning: Learners can gain a better understanding of how vocabulary and grammar work in real-life situations, which can help with retention and practical use of the language.
  2. Reduced stress: The focus on understanding and communication, rather than rote memorization or perfect grammar, can make the learning process more enjoyable and less stressful.
  3. Improved listening and reading skills: By exposing learners to authentic language materials, such as videos, podcasts, or books, they can develop better listening and reading comprehension skills.
  4. Natural language acquisition: As learners are exposed to language in context, they can acquire linguistic structures and vocabulary in a more organic way, similar to how children learn their first language.

Nothing against Stephen Krashen. However, I have yet to meet a German person who spoke German like"do you understand, he has two big eyes. He has black hair." Even the small kids were talking with each other; it was like they were speaking at C3 level (I don’t know if it was -C3 or +C3, LOL). But it was some tough shiit to understand. Very tough. If their conversations were used in the official exam, I would fail it.

Agreed, though. It is better to pick up easy readers rather than a humungous novel from the 19th century if you are starting from scratch. It is better to gradually work your way up in difficulty. That’s why Refold Guide is a great help to consult for and choose content based on gradual difficulty.

Even though I have an itching to pick up a classic of Leo Tolstoy. However, I have delayed it and stuck to reading novels by modern authors and it is turning out to be a great decision when it comes to the language used. At an advanced stage when I have gained a huge deal of vocabulary then I can go for reading old classics. Right now I have numbed my internal itching.


I’ve had the same experience. However, we have to remember that those small kids (who were probably like 4-5 year-olds, right?) have like 10-15k hours (minimum) of perfect input under their belt. That’s the part we keep missing - for some weird reason, we expect to be on that level after 1/5 (or less) of the time it really took those kids to reach their level.

I can see why we have this illusion: Krashen isn’t about to go around telling people that they need 15k hours of input before being able to produce at the level of a native 5 year-old.

Nobody wants to hear that, and if he did say it, we’d probably say that he has ‘extreme’ views about language learning. He’d be even more marginalized by the community than he already is. However, that is the reality, the proof is in exactly what you and I (with our 2-3k hours) have observed when listening to those kids.

This is why comparisons with natives is so unfair on learners - it’s not even close to a fair comparison to make. It’s also why I always raise a smile at people citing those FSI numbers as an estimate to how long it takes to reach ‘fluency.’ Those numbers are based on ‘basic fluency’ which is waaaaaaaay weaker than real, native fluency. The difference between B2 and C2 basically, which is massive and takes a very, very long time to bridge.

And as you say, natives have their own level altogether, that C3+ level. They’ve earned exlusive rights to that level through multiple TENS of thousands of hours of full time immersion.


A drunk Scottish I met when I was working in England had C5+ level for my ears. :DDD

PS: some tiny Italian babies are still incomprehensible to myself when they talk. I usually nod, agree and look at their mothers that speak baby language.

1 Like

I can relate, I’m a native English speaker, from England, and I once had to turn on subtitles when I realised, after about 20 seconds, that the Scottish movie I was watching wasn’t going to be understood without them. :grin:

I suspect every language has almost incomprehensible accents for those outside of the culture. As a Spanish learner, I think it’s el accento anduluz for Castellano. I’m sure there are areas of Italy where it’s the same?

Haha, some parents require the older sibling to translate.

1 Like

Absolutely, in Italy there are tons of dialects that I wouldn’t understand for sure. :smiley:

1 Like

The thing I would point out before you get defeatist is that children have awful study habits. Adults are able to engage in intentional learning and dedicated practice, and get timely feedback from native speakers to dramatically reduce the time it takes to acquire language.

We learn best by briefly and intensely learning about a topic, say 1 “Pomodoro block” and then letting it rest and allowing our brain to process that information. Then revisiting the topic after a day or so.

We also learn well when not trying to learn at all, by simply engaging with and enjoying the thing, but this is not efficient. This is why anyone that has played video games can pick up nearly any game and begin playing, but my wife cannot walk and look at the same time in Minecraft.

The state department and FSI in the US are experts in what they do. They are among the best in the world at getting people to a 4 or a 4+ on the ILR scale (I honestly do not know why we persist on the CEFR when talking about FSI numbers, but I digress). A 4/4+ correlates to a C2 on the CEFR. Maybe not in the matrix description, but the tests the state department does require full professional fluency comparable to that of an educated native speaker. A “C3” or C2+ or whatever you want to say correlates to a 5/5+ on the ILR.

To get those numbers take the FSI estimates, double them once as you are supposed to for out of classroom practice and input, then double them again for “beyond C2”.

This would put category 1 languages at about 2K hours of time using effective methods and actively trying to improve. The key in there though is this is not primarily relying on the unstructured input to make you better, but 2K hours of dedicated practice. Unstructured learning will take you not hundreds and thousands of hours, but 10s of thousands of hours, an entire lifetime for most.


Since I don’t believe that languages should be (or even can be) learned through ‘study’, I’ll continue to believe in the post I made. As I don’t personally believe that we can - as adult learners - “catch up” to the level of that of a native speaker, I’m definitely not being ‘defeatist.’ The battle was already lost the moment I was born and raised in a different country to one that speaks my target language.

This is my personal opinion. It doesn’t mean I’m right, but I do very strongly believe in what I’m saying. You may be right, but I’ve yet to see evidence of an adult learner able to reach the level of dexterity of that of a native speaker.

Sometimes it can be hard to see it at first, but even the very best second language (learned as an adult) speakers are quite some distance from natives. You can see that in vocabularly alone (I’m not talking the use of more eloquent sounding words, which are often a translation from the speakers native language).

There’s a YouTube guy, who is generally considered one of the best speakers of Japanese as a second language there is, and even he admits that his Japanese is nowhere close to his English, and he’s spoken many times about how he lacks sooooooooooooo many less frequent words that natives know.

This guy was a slave to Anki BTW, so although he used immersion as his primary learning method, he absolutely ‘studied’ too, for like 2-3 hours/day, every day, for like 4-5 years. The one thing he didn’t have was 10s of thousands of hours of exposure.


hellion, you say:

“The battle was already lost the moment I was born and raised in a different country to one that speaks my target language.”

I’m sorry to hear that!

But just because you feel this is true for you, it doesn’t mean it is true for everyone else :slight_smile:

I have personal experience of people who have become as good as native speakers, indeed better, after admittedly many years of practice.

Your experience fortunately is not necessarily shared by everyone else.


That is on its face defeatist. I can name several people I know personally that have learned English to that of a native speaker, and far beyond their native language. (In fact, their native languages have atrophied considerably most of the time.)

I do not like discussing language acquisition as a skill, because it often ignores the social and cultural components of how we communicate, and the general knowledge that any native speaker will have. However, language is absolutely a skill that is learned the way we learn everything else. To acquire skills we need:

  • Many attempts with feedback
  • Timely feedback
  • Deliberate practice
    When you think about kids babbling, they get feedback constantly on whether they are following the implicit rules and being understood. (I probably set back my friends’ children years in their language development when I pronounce the “th” sound “d” like they do.)

In chess, you will fall for the same trap a thousand times until you remember it. In language, you will use the wrong preposition or case a million times until you remember it. The key is - is someone telling you it is “wrong”.

On timely feedback, being mad at my dog because she pooped in the house is not helpful when that feedback is given hours later. The feedback needs to be tied directly to the action, or all you have is a dog that has no idea why you are upset. This is why it can be helpful to do Anki or SRS after reading something, or answer questions like a test.

Deliberate practice is something I think you misunderstand. Deliberate practice is when you are putting yourself slightly outside your comfort zone. Learning requires you to be constantly inching up the difficulty into a growth zone. With language there are so many interlocking mechanisms that need practice. You are taking oscillating pressure waves that vibrate little bones in holes in the side of your head, which collide with little nerve endings deep in those holes that convert the mechanical action into electrical impulses. Your brain then interprets those electrical impulses into meaning, decides how to respond, encodes a message and instructs wet air bags in your chest to send air in a precise pattern out different holes in your face while vibrating things on the way and simultaneously moving a large boneless muscle. That is so much stuff, and a deficiency in any area will immediately be noticed by a native speaker. And this doesn’t even touch on reading and writing!

Drilling Anki for 2-4 hours a day is not necessarily deliberate practice, nor is it necessarily effective. You need to break off each one of the requirements and be uncomfortable in short increments for a long time. It takes your brain time to process information and form neural networks. How long? Well, there are two components – duration and time spent in deliberate practice. How much time you get in active study a day without killing yourself will decrease the duration needed, but it is not possible to cram.

Is it possible to “catch up” to a native speaker? Well most native speakers do not deliberately practice their native languages as adults and it is nearly impossible to make them uncomfortable (unless you expose them to Scots). So the answer is almost certainly, look at immigrants that moved and learned the lingua franca. However, the entire premise to this question is ego driven. Why are we comparing ourselves to anyone else? Especially to people in an abstract sense? I personally am only in competition with myself in what I want to get better at. Why do I do anything? Because it is important to me.