Within the umbrella of Comprehensible Input there are different theories of how to use it in language learning. For example, one of the most popular sites for learning Spanish with CI is Spanish Input. The creators of Spanish Input have based it on listening, and only listening with no other learning aids. Do not look up translations in your native language. You listen to the appropriate level, super beginner, beginner, intermediate, whatever, and listen until you understand, then move to the net level. Do not even try to read until you have so many hundreds of hours and the last thing you do is try to start speaking. On Dreaming Spanish their theory is well documented, in fact they explain their method far better than anything I can find here in LingQ.
In LingQ, in sentence mode, we have the translation in our native language. So if we do not understand the sentence, or not sure if our meaning is correct we can check the translation. I see two possible problems with this: One, you may be using course content that is actually more advanced than your current level, and simply using the translation to understand it. And that does not sound like language acquisition to me. The second thing is that one goal of becoming fluent is to stop translating in your head to and from your native language. And for those of us who first spent time with traditional learning methods, it is hard to stop translating everything and start acquiring. So, using the translate to your native language may hinder the natural acquisition of your target language.
I am curious to hear what others think about this, but would love to read something from Steve or other experts about this.
Avoiding to translate stuff you read or hear in your head is a good advice in general, although you have to make some restrictions here. You will not be able to understand anything if the amount of informations unknown to you is too high. So if there are too many unknown words, phrases or idioms respectively or if grammatical structures are used that you don’t know well enough, you will have to at least partially translate what you read or hear, may it be some words only, bigger parts of the sentence or even the whole thing. If you have to do this very often it might be a good indicator that the content you are using is a bit too difficult. However, depending on the language you are learning you may have nothing better.
I personally don’t use translations and I agree that always counterchecking your personal understanding with a translation is counterproductive, as you may rely too much on it, with the effect you mentioned. Only rarely do I translate a sentence (with google translator) if I can not get my head around what is meant. I try to understand how the sentence generates the meaning suggest by google translator, though.
Traditional learning methods usually includes a good mix of input and output both in written and spoken form. Translations are only one part of it, and not necessarely a bad one. Although I’d say that it shouldn’t be the major part of language learning, creating translations in both directions, so to and from the target language, can help to strengthen the already learned knowledge and help getting a better understanding of the details of a language, like the nuancial differences in different expressions or grammar points that might be chosen dependent on the type of content or the context a topic is handled in. Maybe not the most important part when you start learning a language, but this might change later on depending on the reasons why you are learning the language.
Stephen Krashen’s theory of Comprehensible Input is debunked science. You don’t become a fluent speaker from i+1 listening and reading alone (i.e. doing extensive listening and reading). This is simple to prove. And this is not even considering if i+1 extensive reading and listening are the fastest ways to learn a language.
I consider extensive reading and listening to be quite inferior methods, in nearly all cases. You are free to do them as you please, and there’s definitely a cult for these techniques online these days. I much prefer “semi-intensive” methods, where you have the opportunity to look up unknown words with a pop-up dictionary or look at the sentence translations like a bilingual text. The reason is that it’s almost impossible to understand every word from context alone, and solely relying on context alone to guess isn’t the most efficient way to understand meaning. I want to become fluent in my languages of choice in my lifetime, not the next. 
(Side note: I find it very strange the Steve Kaufmann promotes Stephen Krashen’s theory of Comprehensible Input yet looking up words in a dictionary and reading translation sentences in Sentence Mode are pivotal methods in his learning.)
There are studies that positively support the effectiveness of bilingual dictionaries on L2 vocab development. Monolingual ones are good too, but pointless as a beginner.
I personally have not found translation or definitions in the L1 have hindered my ability to transition from translating in my head to thinking in my L2s. Over time, you’ll know most of the words in your input, so you’re knowledge of them will be more and more informed by context through repeated exposure in the L2, and less and less by the original L1 definition. I’ll differ from my friend nfera, as I think this is the primary benefit of extensive reading. (though I mean extensive reading using lingQ, so perhaps that is his semi-intensive reading, in which case we do agree)
you may be using course content that is actually more advanced than your current level, and simply using the translation to understand it.
I understand this concern. To ensure it doesn’t stop with understanding only through translation, you can later re-read and re-listen without sentence mode or lookups. There is some content I re-listened to 10x+ this way over days. At the beginning, I think some great progress can be made plowing through content like this. Over time I repeated content less and less as my comprehension improved.
Thank you for your insight, much appreciated.
Interesting, I have never heard that Krashen’s theory is debunked. But admittedly I have never read about his theory. There certainly seems to be a lot of people claiming success with CI, but they have put i thousands of hours listening. But I moved to Colombia not knowing Spanish, I did not have luxury of only listening for thousands of hours, I had to learn to speak right away, so for me the theory of “listening only” is a no go. But since I started incorporating CI as my main practice, I feel like I am making progress, but it is slow. My goal is to be able to understand my Colombian friends, and I am no where near that yet. Thank you for sharing.
Thank you very much for sharing your insight, it’s encouraging.
In Chinese as a native English speaker, I’d be completely lost without a pop-up dictionary. A new symbol is a (nearly) complete mystery. There’s no reliable way to “sound it out” and guess at the meaning based on Latin/Greek roots or anything like that. Every single new symbol requires a dictionary lookup. Pause 2 sec to think about it and move on. Eventually, you just internalize them and they stick. There are maybe 5% that are stubborn, but that solves itself over time.
Fundamentally, I’d be completely dead in the water without translations.
Anything else that makes the language comprehensible will help. Drawing one of the pictures is just an example for the purpose of demonstration., knowledge of the topic, pictures, translations might be???, etc.Watch the whole video till the end, it is a short video.
Stephen Krashen on Language Acquisition
Basically, he is simply repeating instances of the same word in different context by asking questions. I can assure you that that’s how every German speaker talked to me being a foreigner in Germany. Repeated my active vocabulary. Just to give you a real example. I visited a driving school, so the driving instructor asked me.
She asked:
Arbeitest du?
I said:
Im Moment arbeite ich nicht.
She repeated: Im Moment arbeiten Sie nicht? (same thought expressed in a different way but with the same vocabulary). That’s how language becomes comprehensible. As a result, I learned a new grammar concept to express the same thought when talking to a person face to face (with formal you).
Stephen Krashen’s theory is extreme. It says you can only learn a language through “comprehensible input” (i.e. i+1 extensive reading or listening). With i+1 extensive reading and listening, you will understand everything and be able to speak fluently. The theory is everything else is completely meaningless for learning a language - speaking, looking words up in a dictionary, reading up on grammar rules, reading a translation of a sentence, using a flashcard to drill vocabulary. Anyone even thinking about this would come to conclusion that this is complete rubbish (but many people don’t think logically about it, as they just blindly believe their cult leader). You don’t even need to go out and search for ways to disprove your theory (like a good scientist should do, but alas clearly wasn’t done by Stephen Krashen) because it seems so obviously rubbish.
It’s all easy to claim success. Especially so on the Internet. I have no doubt that you can reach a high level of understanding based on reading and listening alone. Though, I am interested in time-efficient techniques. Language learning is already slow enough due to the sheer amount you need to learn (hundreds of thousands of components to be truly native-like fluent) without handicapping yourself with a slow technique. Furthermore, I’d like to be able to speak and Stephen Krashen’s Input Hypothesis is obviously wrong in that I can become a perfect, fluent speaker without speaking one word.
Extensive reading and extensive listening definitely have their place in the toolbox of techniques. For instance, I extensively listen to some podcasts and YouTube videos in Italian to familiarise myself with different accents, further drill in words I already know, and just overall enjoy a good listen. However, this is not a great way to build vocabulary because unknown words often aren’t easy to guess from context alone. So it depends on what exactly is holding you back from understanding your Colombian friends, whether it’s vocabulary, pronunciation, speed of speaking, etc. as to what exact technique you choose to use. There are many techniques in language learning and you need to change which technique you’re using to adapt to your goals and weaknesses. There is no one technique which will get you to full fluency, and especially not fast.
It’s strange how this person switched from “du” to “Sie” with you. Maybe her first use of “du” was unintentional and she later corrected herself? It does seem a bit odd her using “du,” considering you are her client. But it’s situation and individual independent, of course.
I find that some of these theories can only get you so far. I do a lot of CI reading/listening, but I also spend at least an hour a day going over grammar and vocabulary in my current target languages. As most people know, this is an advantage that older kids and adults have over babies and toddlers learning the language that is spoken at home.
Sometimes you just can’t extrapolate the correct meaning of a new word from context – though I guess that young kids could ask an older person what a given word means even if they can’t look it up themselves in a dictionary. Many kids are great at asking questions.
To me, it’s not a job. I don’t mind a less theoretically efficient approach as long as it’s effective and enjoyable.
Anki might be faster, for example, but it’s boring and I’d just wind up quitting or, at best, doing very little of it. Reading/listening, however, is interesting and engaging so I do a lot of it and don’t feel like it costs me anything at all.
When I started using CI extensively I gave up my verb conjugation charts, vocab lists, rules memorization, any rote memorization, it was such a relief of stress. Now I just look up words and phrases as necessary, practice talking, and so forth, so much more enjoyable and I am learning just as much , or more.
Thank you for sharing such a detailed response. I think with understanding the locals here, it is the speed in which they talk, and accent, local slang. This is why I try to listen to Colombian and Venezuelan sources whenever possible because they sound so much more like the people where I live.
I’ve just started using the translations in sentence mode, and I am finding them very helpful. The translated sentences come out much smoother than the choppy randomness I’m getting from my word-for-word Russian to English lookups. Unscrambling Russian grammar into English takes a bit of getting used to. Sometimes, even when I recognize every word, the Russian words don’t fit together naturally yet. After looking up the vocabulary, I read the translated sentence, and then, when I go back to the original, it starts turning into comprehensible input.
Tolstoy writes some very long sentences! I think I’m seeing the polished work of a professional translator for War and Peace, which is making the content much more comprehensible.
"Stephen Krashen’s theory of Comprehensible Input is debunked science. "
It’s valid.
Wikipedia states:
" * The input hypothesis. This states that learners progress in their knowledge of the language when they comprehend language input that is slightly more advanced than their current level. Krashen called this level of input “i+1”, where “i” is the learner’s interlanguage and “+1” is the next stage of language acquisition."
"You don’t become a fluent speaker from i+1 listening and reading alone (i.e. doing extensive listening and reading). "
You jumped to “become a fluent speaker”. That’s an extreme case.
Comprehensible input (CI) helps immmensely to acquire languages. Bill Van Pattern, Henshaw at al. support this.
The debate is whether you ony need CI and really, at what level you need it.
Paul Nation says you need other methods.
But CI is not “debuned science”.
You say "Extensive reading and extensive listening definitely have their place in the toolbox of techniques. " and I agree.
I use Lingq only and extensively for Greek.
I approximate CI using sentence mode in the sense that I haven’t read of translations being recommended. But then, it’s a very practical choice.
There is almost no Comprehensible Input (CI) for Greek available in terms of a teacher using gestures, stories etc. at a level I can understand.
With Lingq, I can choose a simpel level I can understand or a more advanced level that i partially understand, quite a bit below the 98% recommended in CI.
Other methods said I knew 2000 words in Greek. Lingq got me to 14000. {both words I recognise rather than fully know).
I notice improvements in hearing, understanding, reading and production every 100,000 words that I read using CI.
I don’t see a closer practical way to learn Greek using a CI-like method.
Have you read the chapter of Stephen Krashen’s book Principles and Practice in Second Language Acquisition on his Second Language Acquisition Theory, where he details his five hypotheses, including the Input Hypothesis? Wikipedia is not a good substitute…
This is a direct quote from Principles and Practice in Second Language Acquisition on page 22:
The final part of the input hypothesis states that speaking fluency cannot be taught directly. Rather, it “emerges” over time, on its own. The best way, and perhaps the only way, to teach speaking, according to this view, is simply to provide comprehensible input.
Furthermore, did you ever listen to Steve Kaufmann’s interview with Stephen Krashen on Steve’s YouTube channel? Stephen Krashen believes that you can get a perfect accent from input alone.
Stephen Krashen does not talk about comprehensible input being simply a technique in a toolbox of many techniques, but rather how we learn languages. Full stop. Stephen Krashen’s theory is that you can become a fluent speaker with a perfect accent from input alone.
If the Input Hypothesis is correct and it’s all that’s required to learn a language, why do others like Paul Nation recommend only 1/4 of one’s time be spent “meaning-focused input,” as he calls it (in Paul Nation’s What do you need to know to learn a foreign language? he never mentions the word “comprehensible input” once)? Stephen Krashen’s theory is that you only need input to become completely fluent in all aspects of the language. (This is not to mention that Paul Nation recommends using a dictionary to look up unknown words, whereas Stephen Krashen wants you to completely rely on context alone.)
Stephen Krashen just takes the language learning techniques of extensive reading (i.e. no dictionary look-ups) and extensive listening, relabels them as “comprehensible input,” and then claims that’s all you need to become fluent in a language.
If you don’t believe that you can reach full spoken fluency with a perfect accent from extensive reading and extensive listening (aka “comprehensible input”) alone, then you don’t believe Stephen Krashen’s Input Hypothesis. It’s a fundamental part of his hypothesis.
I’ve mentioned this before, but to see if you can reach full spoken fluency with a perfect accent from “comprehensible input” alone, you can look at certain populations which are under such circumstances for decades. An example of this are Austrians and Swiss Germans in various areas. They watch television and movies in Standard High German, but never speak it. If you go to these populations and get them to speak Standard High German, will they speak it fluently with a perfect accent? The answer is no. Stephen Krashen could’ve went out and tried to find these populations (of which there would be several) to test his Input Hypothesis before he published it, but he clearly didn’t.
TL;DR Stephen Krashen’s Input Hypothesis is so clearly and obviously wrong.
I have found that when using the sentence translations I do discover phrases consisting of two or three words that taken separately do not make sense, but as a phrase take on a new meaning. In this way the sentence translation is very helpful.