Useful info about CI methods?

[leosmith]

I got negative 46 votes in a forum for merely stating “CI is not a skill”. I was told that I was being pedantic, and not contributing anything useful to the discussion. Some foul language was used, but none of it from me. Anyway here I am, in a better forum, trying to share some useful information about these so-called comprehensible input methods, which I will to refer to as CIMs. For the uninitiated, CIMs tend to have the following in common:

1. They encourage the consumption of lots of CI (hence the name), meaning reading and listening to things that are somewhat understandable. The input is often graduated, and it is considered optimal if the learner consumes items slightly above their current level (i+1 for Krashenites).
2. Except for consuming CI, they discourage all other types of learning activities for a longer period of time than most language learning methods do. In other words, there is a “silent period” while consuming CI, during which the learner is to avoid conversing, writing, studying grammar, etc.

Why am I writing this post? Because CIMs are the latest fad in language learning methods, and it’s hard to hang out in any of my favorite sites without seeing CIM fanatics giving advice to/trying to recruit learners. They make erroneous statements like CI is a method, skill or immersion. Sometimes I understand what they mean, sometimes I don’t. This is a bit annoying, so I sometimes correct them even if it’s not absolutely necessary. But they also often make claims about the benefits of their method that are questionable.

Fake benefits
• The method yields superior results.
• You never need to study grammar explicitly.
• After the silent period the learner has superb pronunciation.
• After the silent period conversation is easily acquired.

I disagree with all of these. I want to focus on the first one though – CIMs don’t give a learner superior results imo. The long silent period, and resistance to doing non CI-consumption activities prevalent in so many of these learners really slows their progress. The data presented many years ago from ALG indicated this. Posts from ALG learners, and my personal interaction with them point to the same conclusion; it’s not a very efficient method.

These days, when I ask CIM followers to state the reason why their method is superior, they typically say something with Krashen logic. It’s natural, like a baby. Acquire, don’t learn. Etc. These are easily refutable. But I’ve heard a slightly different reason that I find interesting. I’ve been told that CIMs work the best because “CI is immersion” and “immersion is required to achieve fluency”. Other methods aren’t immersion, so they come up short. Yikes. Even though it seemed intuitive, I had to really dig deep to explain why this wasn’t true.

Obviously, CI is not immersion; it’s just input. But let’s assume they meant “consuming CI is a form of being immersed”. When I asked what they meant by immersion, I was told “Immersion is defined as deep mental involvement in something”. The problem with this definition is that it’s such a low bar, and it’s not specific to language learning. When you think about it, any kind of studying at all, for any amount of time, is immersion by this definition, so CIMs don’t distinguish themselves here. A better definition is “Immersion in language learning refers to the process of surrounding oneself with the target language in an environment where it is predominantly spoken. This approach aims to simulate the natural way in which individuals acquire their native language by exposing learners to authentic language use in real-life contexts”. A CIM follower may be immersed per the second definition, but I don’t see the method as being the reason for the immersion. It’s an individual choice.

Regarding “immersion is required to achieve fluency”, of course, deep mental involvement is required, but one does not have to go to the extent specified in the second definition of immersion above. I don’t like their willy nilly use of immersion – it’s reserved for super dedicated language learners imo.

True Benefits
• The CIM pay sites develop lots of CI for their followers, much of it free for all to use. This is fantastic!
• CIMs make it very clear that one needs to consume tons of CI to reach a decent level. Many methods don’t do this. I suspect this is one reason why CIM followers experience better progress than with previous methods, and the progress generates some of the hype.
• CIMs also give learners who have good reasons for delaying production, grammar study, etc. legitimacy. Maybe they are shy, or they absolutely hate grammar or something, and are tired of being made to feel inferior. These learners can thrive in a CIM.

Do you have any useful information about CIMs? What advantages/disadvantages do they have over other methods?

[rdearman]

Do you have any useful information about CIMs? What advantages/disadvantages do they have over other methods?

I have come to the conclusion that I+1 = 0

CI seems to be just another fanatical dogma. I succeeded after a couple of Super Challenges, but I didn't pick things that were I+1, I read and watched anything I could get my hands on, after having done lots of textbooks and lessons. I also did hundreds, if not thousands of language exchanges. So no silent period for me!

Dogma common indicators:

Rigid Belief System: Dogma often involves a set of beliefs that are fixed and not open to questioning or revision. This rigidity can lead to resistance to new ideas or evidence that contradicts the established beliefs.

Authority Over Inquiry: Dogmatic systems typically place authority figures or texts above individual inquiry or critical thinking. Questioning the beliefs or teachings is discouraged or even punished.

Absolutism: Dogma tends to present its beliefs as absolute truths, leaving little room for nuance or differing perspectives. This can lead to a black-and-white view of the world, where everything is seen as either right or wrong, good or evil.

Closed-mindedness: Those adhering to dogma may exhibit closed-mindedness towards alternative viewpoints or evidence that challenges their beliefs. They may reject dissenting opinions without consideration.

Inflexibility: Dogma often resists change and adaptation. It may be resistant to updating its beliefs or practices in response to new information, societal changes, or evolving moral standards.

Exclusivity: Dogmatic beliefs may lead to an us-versus-them mentality, where adherents see themselves as the sole possessors of truth and view outsiders as misguided or inferior.

Emotional Attachment: Dogma can create strong emotional attachments to beliefs, leading adherents to defend them fervently, sometimes even in the face of contradictory evidence or logical arguments.

Imposition on Others: In some cases, dogmatic beliefs may be imposed on others, either through social pressure, legal means, or even coercion, leading to restrictions on individual freedoms or rights.

[zjfict]

It seems CI has devolved into a cult in places. Folks take a neglected and important aspect of SLA in times past and fanaticize (is that a word) it. Sad.

[Querneus]

True Benefits
• The CIM pay sites develop lots of CI for their followers, much of it free for all to use. This is fantastic!
• CIMs make it very clear that one needs to consume tons of CI to reach a decent level. Many methods don’t do this. I suspect this is one reason why CIM followers experience better progress than with previous methods, and the progress generates some of the hype.
• CIMs also give learners who have good reasons for delaying production, grammar study, etc. legitimacy. Maybe they are shy, or they absolutely hate grammar or something, and are tired of being made to feel inferior. These learners can thrive in a CIM.

Very good post, leosmith, thank you. Especially this list of true benefits.

I dislike how I've often gotten pushback from language teachers for saying that different methods likely work better for different people. It is things like that third benefit that I have in mind. Personality, and in particular, the activities they like/dislike seem to me to be things that matter.

I have come to the conclusion that I+1 = 0

Hmm. Does i + 1 = 2/(i + 1)? Seems unintuitive and wrong, but I'll leave it to people who're better at math than I am.

CI seems to be just another fanatical dogma. I succeeded after a couple of Super Challenges, but I didn't pick things that were I+1, I read and watched anything I could get my hands on, after having done lots of textbooks and lessons. I also did hundreds, if not thousands of language exchanges. So no silent period for me!

It sounds like you simply didn't try CI though. Which is fine? Lots of people don't. I haven't.

[Saim]

In my experience if you focus on comprehension too much you end up with lots of gaps of all kinds, even if you're doing dozens and dozens of hours non-stop (cue the Krashenite — "many dozens of hours is nothing! you should be doing hundreds! thousands! millions of hours of Romanian-language anime as a working adult!"). You can definitely learn to communicate in some sort of way with some sort of largely CI methodology, but it generally leads to way more gaps in grammar, idiomaticity and active vocabulary than most CI proponents will recognise.

Personally, I find that doing various output and recall activities — speaking (including both interaction and self-talk), cloze exercises*, translation exercises and rote memorisation of vocabulary — not only directly makes my output better, but makes me actually pick up a lot more from input, making my output again better indirectly. Personally I believe that this gives a lot of credence to the noticing hypothesis; i.e. that you actually need to prime your mind to pick things up from input. Not to mention that this all helps you develop more intrinsic motivation because the different activities reinforce each other, helping you derive more enjoyment from the process.

This also reminds me of the distinction our own Cainntear has drawn between superficial activities and the underlying mental processes they are supposed to stimulate. I think that if you avoid output and recall activities you can end up "going through the motions" in your input and consuming for the gist rather than actually paying much attention to the way things are actually expressed in the language. And if you don't notice the specific words and collocations used, it's unlikely you'll develop much idiomaticity or automaticity in your speech.

*A lot of my cloze cards in Anki are effectively "i+0" — I know all the words in the sentence well but want to remember the exact collocation.

[leosmith]

I have come to the conclusion that I+1 = 0

Unlike you, I prefer to read "level appropriate" material. However, 2 points:
1) It's not necessarily the most important factor to me. For example, interest trumps it in many cases.
2) Since I started using a reading tool, I have greatly expanded what fits into the "level appropriate" window.

[Dragon27]

Does i + 1 = 2/(i + 1)?

What's this? Looks wrong to me. Here are some correct alternatives:
i + 1 = 2i/(i + 1)
i + 1 = 2/(1 - i)
We're talking about the complex numbers, right?

[GawainStan]

I can't speak for the efficacy of CI in general but I often wonder if the biggest advocates of the system are people who only ever learn 'popular' or dare I say, eurocentric languages. I have yet to find the Persian equivalent to Dreaming Spanish or a vast array of n+1 audiovisual material covering Khmer, Telugu or heck even Bengali.

I feel that very often the big method of the day, whatever it may be, gets popular because the languages most anglophone people learn are relatively easy enough that any method will eventually teach you enough of the language. Even the dread Rosetta Stone isn't really such a bad method for learning romance languages, its flaws become evident only in 'Asian' and Slavic languages (from what I've heard).

Of course this is just my opinion and I could be wrong. I just personally haven't benefitted from CI except in Latin (if Lingua Latina per se Illustrata counts as CI proper) because many of the languages I'm interested in simply don't have CI-focused resources for them. I will of course acknowledge that there's a dearth of good resources for some languages in any format simply because fewer people study them.

[rdearman]

I don't think it is possible to "try CI", since it isn't very well-defined. For example, I'm learning Korean. I know the sum total for SFA in Korean. So what is 0+1? But more importantly, how can I, a person who cannot read Korean, determine if a book is 0+1 ? I can't, the only way I could find that out is to get a Korean teacher, and they feed me texts which are 0+1 and if I have a teacher, why would I need CI? I might decide that just because the book is in the format of a comic book, it must be 0+1, but if you've ever read any of the French BD (comics) then you'll know that isn't the case! But let us assume that I, a novice language learner, has enough knowledge to work out which books are 0+1. CI alone may not provide enough explicit instruction on grammar rules, vocabulary, and other aspects of language structure. This could lead to gaps in my understanding and hinder overall proficiency.

It is likely that CI may lead to slower progress compared to more structured approaches that provide explicit instruction and practice opportunities. Also, CI relies heavily on context to make meaning comprehensible, which may not always be available or clear, especially in authentic language use situations or with unfamiliar topics. Again, as a 0+1 beginner, I wouldn't be able to determine this myself.

There is a huge potential for misinterpretation. I may misinterpret or misunderstand input, especially if it's too complex or lacks appropriate scaffolding. Of course the CI priest will tell me I shouldn't be reading complex things, but I can't know this because I don't yet know the language. This leads to confusion and frustration with the language and a good chance I'll quit. Not all input provided may be comprehensible or appropriate for the learner's level and needs. In fact, I'd say very little input will be appropriate if you're self-selecting the input. But you can't tell CI practitioners this, because it is like trying to nail molasses to a tree.

CI emphasizes understanding, it may not provide enough opportunities for learners to practice producing language, which is essential for developing speaking and writing skills. People do not magically start speaking a language after X number of hours reading and watching TV. You get better at what you do, so if you're just doing input, you get fantastic at understanding input.

While CI can help learners understand language input, it may not adequately prepare them for real-life communication situations where they need to produce language spontaneously and interactively.

Unlike you, I prefer to read "level appropriate" material.

How do you determine this? Also, I have followed your logs and posting both here and on HTLAL, and you're using textbooks, audio courses, online tutors, you do conversation exchanges, you also write in your languages. So ... I don't think you're the poster boy for the CIM crowd.

[emk]

1. They encourage the consumption of lots of CI (hence the name), meaning reading and listening to things that are somewhat understandable. The input is often graduated, and it is considered optimal if the learner consumes items slightly above their current level (i+1 for Krashenites).
2. Except for CI, they discourage all other types of learning activities for a longer period of time than most language learning methods do. In other words, there is a “silent period” while consuming CI, during which the learner is to avoid conversing, writing, studying grammar, etc.

Why am I writing this post? Because CIMs are the latest fad in language learning methods, and it’s hard to hang out in any of my favorite sites without seeing CIM fanatics giving advice to/trying to recruit learners.

"Comprehensible input" isn't truly new. For a similar idea from well before Krashen, the direct method was invented around 1900. It mixes CI with echoing and some active exercises. Here's an example Spanish textbook from 1922, which is clearly trying to do comprehensible input (and taking advantage of Latin cognates to do it). We keep reinventing the 6 same language learning methods on a 2-generation cycle, in response to shortcomings of the previous fad.

But after the "direct method", we see the first editions of Assimil around 1929. These underwent some tweaks before setting on the current design, which features:

1. L2 audio and text, delivered in progressive daily lessons.
2. L1 text, which the listener is encouraged to use as a comprehension aid, and to then cover up once the L2 audio is comprehensible on its own.
3. Footnotes which point out interesting details, and an end-of-week review that summarizes the most important grammar the student has recently seen.
4. A 50-day delay before producing any output more serious than a few half-hearted "fill in the blank" exercises.

And Assimil works fine—it's a consistent and reliable course. Admittedly, the course format is simply too short for something like L'Égyptien hiéroglyphique, where the student should probably "overlean" the lessons to get the most out of limited text. And well after Krashen's theories became a trend, we see the popular Lingua Latina Per Se Illustrata (1995).

In terms of online language-learning communities dedicated to input, we have the venerable Antimoon, which argued for Krashen's ideas. And of course there was AJATT, which succeeed for some learners—but not all. Tons of people have used methods built around generous input and a bare minimum of grammatical instruction. And these methods do reliably produce passive skills.

Fake benefits
...
• You never need to study grammar explicitly.

I think it's worth studying some grammar at some point. For example, I picked up my first actual French grammar book about 2 weeks before my B2 exam, and skimmed it in about 3 hours. But that time was extremely well spent. It was helpful to see the actual rules underlying things I'd been using for a long time. And of course I did grammar study before that. I made a half-hearted attempt to read Assimil's weekly grammar reviews, and I would definitely Google strange constructions I'd seen. But mostly I read, and watched, and spoke, and wrote at least 1,500 words (and got them corrected). And I had at least 15 sessions with a good tutor who'd point out weird phrasing.

But at the same time, you cannot reach C2 using grammar workbooks. As I've previously mentioned, I have a delightful 1,800-page descriptive grammar of standard written English. Native speakers who read enough books know basically everything in it. And non-native speakers who've been using English for 10+ years know a lot of it. Nobody teaches most of the things in this grammar to ESL students. In comparison, the best C1 grammar workbook I've seen is Grammaire progressive du francais: Niveau perfectionnement. But this is a slim little thing, containing only a tiny fraction of the information that would be needed for a true linguist's grammar of French.

Still, I never did an FSI drill, and I never did a single page of a grammar workbook. At least some students can get away with very small amounts of "traditional" grammar study.

• After the silent period the learner has superb pronunciation.

Adults only get "superb" pronunication if they're naturally gifted, or if they put in a lot of work. The are no guarantees of "native" pronunciation for languages learned after age 6.

• After the silent period conversation is easily acquired.

I really doubt this. If anyone was following AJATT/Silver Spoon/Neutrino back in the day, Khatzumoto made a herioc effort to teach people using 100% input. And while many of his customers reached fairly good levels of comprehension, many of them were completely unable to speak. Khatzumoto ended up refunding a lot of money.

At least in my experience, passive languages can be activated by using them, but "easily" is definitely not a word I'd use here. My description of the process was, "I feel like my brain is going to leak out my ears."

I disagree with all of these. I want to focus on the first one though – CIMs don’t give a learner superior results imo. The long silent period, and resistance to doing non CI-consumption activities prevalent in so many of these learners really slows their progress.

I think an Assimil-sized "silent period" of 50 days is perfectly reasonable. As a heavy Anki user, I find that 1- or 5-day-old comprehension cards can be difficult. But 30-day-old comprehension cards are often almost effortless. I don't know why this is. But I'm happy to push output back until I reach that milestone. I find the idea of learning a new construct and using it the same day to be slightly masochistic.

I am studying Spanish with extremely input-heavy methods, though I've recently introduced something like Assimil's active wave. And I'm finally using Anki to learn verb endings. But I could follow along with 30–40% of a specific TV should before I could actually conjugate the present tense of ser. And overall? I'd say I'm making decent progress, especially relative to the time I've put in. So it's clearly possible to rely on input-heavy methods and tiny amounts of grammar study, without sacrificing results.