4 Likes
Well worth watching. In my opinion, comprehensible input, in the sense of Krashenâs theory, is a con job. As far as I can tell, it was never widely accepted in linguistics, and there is now plenty of proof that contradicts its tenets. It has clearly become very popular in some circles, especially among YouTubers. I wonder who the original social media promoters of Krashenism were?
3 Likes
I mean this is LingQ? Isnât this CI?
1 Like
Ha! Some people say yes, and some people say no.
Pretty much everyone recognizes input is important. The âCIâ in the video refers to Krashenâs position that ONLY input matters in language acquisiution. The research LoĂŻs shares in the video are just some other researchers positions that highlight other things that are important, in addition to input.
3 Likes
The term Comprehensible Input (CI) was coined by Professor Stephen Krashen, as part of his input theory. According to his method, we learn through input alone, we cannot learn through explicit study, we always learn in a set predetermined manner plus a couple more hypotheses. He states that output does not improve the learnerâs competence in the target language. Itâs hokum.
Some people take CI as the idea that we benefit from graded input, which of course we do. But thatâs not Krashenâs theory. Anyone with any sense can see that graded input is beneficial, when it forms part of a broad learning methodology.
So no, LingQ is not CI per se, but you can use it if you want to follow Krashenâs nonsense.
4 Likes
I like the simple good old Practice makes perfect method.
After some studying it is the practice in reading, writing, listening, speaking that gets us good in those areas. Practice a lot by consuming and using the language.
4 Likes
Vegetables are good for you but eating nothing but vegetables wonât help. As with everything in life, balance and moderation are important. Listening, reading, speaking, and writing are the four âfood groupsâ of language learning. All are equally important and reinforce one another.
5 Likes
In my opinion there are no arguments against comprehensible input being great but its another story about how much CI you need to be able to start speaking naturally without tons of practice. If it IS possible then i think people underestimate the amount you need to actually pull that off. There are Japanese learners who have said they have done this and either they are telling the truth or they practiced speaking more than they let on. In my case with about 2k-3k hours of Japanese study, i understand way more than i can produce on my own. Its frustrating but for me i feel like speaking is something that needs to be practiced separately but perhaps im still too early to naturally output without trying. Also everyone gets different levels of CI so its hard to compare yourself to others. In a perfect world where everything was perfectly comprehensible then that may be a different story
3 Likes
I see CI or input in general as like the bricks, but the study/output as the mortar that holds the bricks together. Itâs foundational but not sufficient.
Luca did a great video where he described the interface theory where Krashen and ALG assumes no interface between conscious and unconscious acquisition.
But science and those of whom seem to have acquired languages the strongest in adulthood point to a âweak interfaceâ where study/practice eventually leads to procedural learning.
I spent nearly a year immersing in Dreaming Spanish and many other CI content for my Spanish and I could not just automatically speak. It took a lot of effort, but i had the materials in my brain to speak well. I discovered how speaking is the skill of constructing sentences without conscious left brained thinking. Similar to predictive AI models, with enough input you let the next thing to express come, and itâs a slow and steady process to learn to speak.
Another missing piece in the CI theory is that itâs missing the social context in which first language is acquired.
Weâre not passive recipients, weâre active relational participants. Especially in the output stages.
I also missed A TON of the structures of Spanish with CI. Despite thousands of hours of CI and immersion. For example I never picked up âa lo large deâ until I made it explicit once in LingQ or google translate. Iâm skeptical that weâll automatically pick up every linguistic structure in a language by only extensive CI immersion.
CI is like the base of the pyramid but is best combined with something that makes it explicit.
6 Likes
Perfect analogy! Thanks!
1 Like
The theory of CI is so alluring: easy to understand method that promises the moon. What I liked about the video was how often studies that actually measured results were cited. Most of the CI stuff never bothers to ask the hard questions necessary to validate the hypothesis.
3 Likes
This is the part that annoys me on your behalf 
, the idea that ânatural output without tryingâ is even a thing. In most cases where people claim it, thereâs more to the story or theyâre selling something.
Consider that some children in multilingual households grow up as receptive bilinguals, able to understand their heritage language, but unable to speak it despite a mountain of CI hours. Their very existence certainly challenges the idea that input alone produces effortless output.
Here is another cautionary tale: https://www.youtube.com/watch?v=uPacj1w-yX4
Good luck and é ĺźľăŁăŚä¸ăăďź
4 Likes
IMO itâs important to note that Krashen speaks of language âacquisitionâ â not learning, not output.
Acquisition is the subconscious, pattern-absorbing process like how kids pick up their first language, while âlearningâ is the conscious, rule-based stuff. And output is output.
Learning and output stabilize and strengthen fluency, but they are not, in Krashenâs framework, how we acquire language.
Krashenâs acquisition is actually similar to the training LLMs undergo when absorbing languages.
1 Like
As noted in the video, there is no reliable way to empirically differentiate between acquisition and learning in a speakerâs language use, making the theoretical foundation shaky.
3 Likes
Perhaps I misunderstood you, but Krashen is not silent on output. He states that output emerges as the result of input (the acquired system) and that this acquired system is responsible for our fluency. We acquire in only one way (understanding). Learning has only one function, and that is as a Monitor, or editor, in his view. This learned knowledge cannot become acquired knowledge.
Based on this framework, all spontaneous language must come from the acquired system, built by input.
From what I have seen him say, he seems to be consistent here. When asked how to improve output, his answer is always some form of âget more inputâ.
Let me know if you think Iâm missing something here.
7 Likes
100%. My case as well. Speaking ability surged once I started speaking everyday (in JP). I think a lot of language learners who are fluent in another language donât really mention the WHOLE story of their progress.
5 Likes
It should be completely obvious that comprehensible input alone is not enough. The only way to learn to produce output, is to produce output and keep doing it. Itâs similar to how when you learn math, you canât just be presented math problems and then be presented the solutions to these math problems over and over again, you actually have to get math problems you have to solve on your own, repeatedly, to get good at math. So basically, you have to repeatedly be tasked with creating your own sentences (forming speech or text from your own thoughts) and get corrected to some degree at least some of the time. Of course you need input before you can produce output, otherwise you have no vocabulary you can use and no idea of sentence structure, so comprehensible input is an essential component of language learning, but not the only thing you need.
12 Likes
With respect, surely this is an oversimplification of Stephen Krashenâs theories? You seem to have a serious gripe about his scholarly work⌠âhokumâ, ânonsenseâ etc.
It is clear that according to Krashen language acquisition PRIMARILY relies on comprehensible input. But Krashen most definitely acknowledges the role of conscious learning through explicit study, and particularly as a âmonitorâ or âeditorâ for output when you get to that stage.
I think most language learners can appreciate that, sometimes mysteriously, language is âacquired subconsciously through exposure to understandable messagesâ, which is his definition of comprehensible input. After a lot of input, particularly through massive reading, you suddenly can manage a phrase and think âwhere did that come fromâŚ?â And the answer, as rokkvi has just indicated, is that this is derived from your input by reading or listening.
Professor Krashen at no time rules out âconscious learningâ by explicit study, as he himself enjoys grammar and indeed even some old-fashioned rote learning. This knowledge, he says, can then be used to âmonitorâ and self-correct your output.
Personally I find Krashen has had some very good ideas, foremost of which is to just keep on reading and listening with massive input, and then adjust with deliberate learning and âmonitoringâ when you come to output for writing or speaking. I can particularly appreciate his oft-told story of the young girl reading Herman Melvilleâs âMoby Dickâ. When Professor Krashen suggests gently that this might be a bit hard (and clearly even beyond his âinput + 1â theory) she responds âOh, I just skip over the difficult words!â That seems to me an inspiring way to âacquireâ competence by reading âcompellingâ material.
5 Likes
Thanks bembe, I always appreciate your posts.
The version of CI and acquisition you outline is one I can get behind: primarily, but not exclusively, input. I also enjoy Krashen (and his anecdotes!).
That said, in his own writing heâs very explicit that conscious learning doesnât become acquisition. It can only function as the Monitor, and the Monitor affects performance, not competence. He even states bluntly: âLearning does not turn into acquisition.â So while he personally enjoys grammar study, heâs clear that he doesnât see it as a pathway to acquisition.
Iâd be curious if youâve come across anything in his work that contradicts this. Many later critics of his model focus exactly here, arguing that interaction, instruction, feedback, or output also contribute to acquisition, which sounds closer to the view youâre describing.
3 Likes
I think what youâre describing is essentially âComprehensible Input.â
2 Likes