4 Myths about Comprehensible Input

greek vaseEver since Stephen Krashen made it central to his Input Hypothesis, since rebranded as the Comprehension Hypothesis, “comprehensible input” (CI) has been part of both academic discourse and shop talk among language teachers. It’s easy to have misunderstandings surrounding a concept that has engendered so much spinoff research and so much water-cooler conversation, so let’s clear up some stuff!


Myth 1: Comprehensible input is a method

Fact: Comprehensible input neither is nor requires a particular method. It’s simply a type of language that occurs either by nature or by intention. If someone says or writes something and someone else who isn’t a native speaker understands, comprehensible input has happened. Many language teachers make an extra effort to provide their students with lots of language that their students can understand, because of the widely documented association of comprehension with the acquisition of language. Some teachers have combined ways of providing and sustaining CI into full-fledged methods, such as Teaching Proficiency through Reading and Storytelling (TPRS®), but comprehensible input itself is not a method, approach, or philosophy.

See also: Ellie Arnold’s “CI Methods: an obvious epiphany


Myth 2: The need for comprehensible input is hotly disputed

Fact: Just about every theory of second language acquisition (SLA) affirms the importance of target-language messages that learners can understand. What’s disputed is whether comprehensible input is all you need in order to acquire a language or whether other factors or actions (social interaction, explicit awareness of textbook rules, targeted practice,…) can or must play a role.

Even if there weren’t support for the value of CI from diverse researchers and practitioners, remember that, since comprehensible input is simply a type of language event, there is not really such a thing as being “for” or “against” it, any more than there is such a thing as being for or against trees. Trees spring up by nature. One can also plant trees. If one thinks that trees are good and that there aren’t enough, one plants more. Teachers convinced of the efficacy of CI for acquisition find and plant lots of CI. Teachers who haven’t heard, haven’t noticed, or aren’t convinced of the efficacy of CI spend more time on other things.


Myth 3: Comprehensible input is inauthentic

Fact: Since comprehensible input is any bit of language that is understood by a particular learner, and there are learners who can understand at least some “authentic” language, CI can be “authentic” even in the narrowest definition of authentic language as “language produced by a native speaker for other native speakers.” What is rare is that Novice and Intermediate learners thoroughly understand extended instances of language (several minutes of speech, several continuous pages of writing) that does not take their proficiency into account when it is created.

In any case, there are alarming problems with avoiding input produced by or for non-native speakers or implying that such input can’t be of great value. Some are philosophical: What is the point of calling anything a native speaker says to a non-native speaker “inauthentic”? What are the implications of labeling all the Spanish I will ever produce until the day I die, no matter how correct and situationally appropriate, “inauthentic”? What does it mean to say or imply that every single thing ever said in a language class is “inauthentic” because it is said either by or for the sake of a non-native speaker? Then there are practical issues: For one thing, good communicators modify their speech all the time in order to be understood better. Should a language classroom be the one place where we avoid this basic feature of human communication? For another, although it is not logically necessary, my observation is that students who are exposed only, or almost only, to “authentic” input tend to receive not only a much lower amount of comprehensible input, but also a lower total amount of target language input of any kind.

So, comprehensible input does not have to be “inauthentic,” and there are really good reasons to pursue it even when it is.

Edit 12 Oct 2018: Check out Eric Herman’s concise and cogent post “Let’s Get Real (Not Authentic)


Myth 4: Comprehensible input is outdated

Fact: By now you know what I’m going to say–how can language that a learner understands be outdated?

That said, one does come across comments that Krashen’s work is outdated, and the question of the value of comprehensible input is sometimes attached to either the intention or the reception of such comments, even though Krashen is far from being the only scholar treating CI. These comments often originate in the UK, where, for various reasons (including, as rumors have it, non-academic ones), Krashen’s influence is not as apparent as elsewhere. In the United States, where I do most of my teaching, training, and observing, reliance on comprehensible input is as strong as ever–probably stronger than ever, as Krashen continues to interact with teachers more closely than almost any other researcher, as his work has had time to be taken up and expanded on by many other researchers, as the results of this research have trickled down to a critical mass of classroom practitioners, and as these practitioners have found their students to be unprecedentedly successful.

Have you heard other misconceptions about comprehensible input, or do you have something you yourself wonder about it?

Posted in Learning, Teaching and tagged , , , , .


  1. Justin, I appreciate you writing about #3. I agree that a preoccupation with “authentic resources” can debilitate a language program and further, distress students. I like how Paul Sandrock and Paul Nation both roughly defined it in a past ACTFL forum conversation, to paraphrase, messages that are communicated among two or more human beings.

    I question this statement: “comprehensible input itself is not a method, approach, or philosophy.” I agree that it should be distinguished from methods (e.g. MovieTalk, TPRS, TPR). But I consider Teaching with CI my underlying approach to teaching Spanish, guided by the philosophy that CI drives acquisition. I’d like to hear your thoughts about this.

    • Hi, Jim, Thanks for sharing Sandrock and Nation’s definition of “authentic,” which is much better than “by and for NSs,” as it makes the issue about the purpose of the utterance–communication–instead of about who says it to whom.

      It’s true that “CI” has become shorthand for an approach or set of approaches, such as “Teaching with CI,” but comprehensible input itself is simply a type of language (“understandable messages”) that (a) can feature, intentionally or not, in a variety of approaches to language instruction, and (b) also occurs outside of instructional settings.

      I’ve been contemplating writing a post about “Teaching with CI”–your interaction is good motivation to go for it! Thanks again, Jim.

  2. I also appreciate your writing on #3. To me all messages are authentic since we use language to get our message across whether or not it is our 1st or 4th language. If a baby can get a message across without full access to all the functions of language, then students do too, and it is authentic.
    As Jim said above, it can distress students, but also teachers who are non-native speakers. The goal it to be understood.

  3. Curious me: #4 – on the question of supposed “out-of-date” research (Krashen). What newer/better research do the naysayers hold to? My observation of “communicative approaches” (especially in Britain) is that they are based on ideas far older than Krashen’s original hypotheses. What have you observed, Justin?

    • Great question, Jody! It varies, but the approaches that seem to predominate in the UK are “skill-based” ones that view language-learning as working the same way as (cognitive) skill development in general: extensive, targeted practice through which linguistic knowledge gradually moves from declarative to procedural memory. These approaches are indeed older than Krashen’s formulation of the Input/Comprehension Hypothesis, but holders of a skill-based position sometimes point out that Krashen’s most prolific period pre-dated certain brain-imaging studies on which they call for support.

  4. The British, almost as well-known in the world as Americans for their language skills, have made teaching English to the rest of the world a national enterprise. For generations they have been creating language schools in the far flung corners of the globe and Oxford University Press and Cambridge University Press produce mountains of textbooks and materials for English language teachers. They are not about to let an American professor tell them how to do it.

  5. Pingback: Beginning of the year reflections – University of Stoa

  6. Terrific post. I’m interested in the implications of Myth 3, and the theoretical underpinnings that support the refutation of this myth. Thank you for sharing your thoughts.

  7. Pingback: What is comprehensible input? | edcampCIVa

    • Hi, Debra, thanks for reading, commenting, and making this connection!

      The “i+1” element of the Comprehension Hypothesis does evoke the concept of ZPD, but the mechanisms proposed by Vygotsky and his heirs are quite different from those proposed by linguists who discuss the role of comprehensible input in acquiring a language. In fact, scholars who have attempted to apply Vygotsky’s work to second language acquisition have often done so as an explicit critique of the Comprehension Hypothesis and its focus on cognition. If you’re interested, I’ll see if I can dig up some sources!

  8. Great post. Love the writing style.

    On #1, I think there’s a complication.

    Isn’t there an argument that if enough people use CI to mean an approach (or group of approaches), to understanding language acquisition, and teaching language, that deeming it not such is arbitrary?

    One may wish it wasn’t used to mean such, and understood as meaning such, and with good reason. But given, I think I can say, that it quite often is, also with good reason, namely that there’s no other widely shared way of saying “thinking about language acquisition or teaching in a way consistent with the concept that people acquire language best or only when they receive CI …”, who can say that it doesn’t mean such?

    Evolution of language, just like “comprise” used to have only one meaning but has come to have two.

    • Thanks for reading and commenting, Jim. You make an excellent point about the semantic shift of the term “comprehensible input/Comprehensible Input,” a shift that seems even more complete in the abbreviation “CI.” As someone who takes a pretty thoroughly descriptive/descriptivist approach to language, I am happy to concede the point, though I would (and do, when the setting seems right) suggest that it is worth making a distinction between comprehensible input as an instance of language that is understood by a particular learner–basically a type of stimulus–and Comprehensible Input as a theory, method, approach, family of approaches, or philosophy. Many are trying to take this distinction seriously with terms such as “CI approaches,” “Comprehension-Based Communicative Language Teaching,” “the CI umbrella,” and “CI-rich teaching.” Have you found any particularly useful ways of clarifying this distinction?

      It’s also worth noting that the term “comprehensible input” is still used in academic literature to mean “language that is understood,” not (to my knowledge) a particular theory of acquisition or way of teaching, and that Krashen himself does not equate it to a teaching method. As you imply, though, my post includes an element of wishful thinking or even lament similar to that in my pair of posts “The Bummer about Acquisition,” and kicking against the goads of crowd-sourced semantic change is likely to yield frustration.

Leave a Reply

Your email address will not be published.