Krashen and "Krashenite"

[Cainntear]

Clearly you are a lot more annoyed about this than I am. I don't really know why.

Because I'm sick of teachers and online self-help gurus parroting his stuff and feeling all please with himself when they do so.

I do know what previous linguistic world Krashen's theories came into and what he was addressing, yet I don't really see what relevance it has to considering the merits of his views. We can mention names and people, but it's best to just say what they say/said and then discuss those as counter-views, support views or whatever

Well, the unspoken point of my rhetorical approach was that lots of people don't know this stuff, and whether you do or not, it's hopefully quite clear to you that a lot of people don't.

For those who aren't familiar with the wider literature, Merrill Swain's response was the "Output hypothesis" -- basically saying that it's all well and good understanding, but that doesn't lead to being understood. If you develop a theory of what the language is, you need to test that theory, and if you're wrong, you'll get feedback and hopefully refine your theory. I don't remember the specifics because I was doing language about 20 years ago, and Merrill Swain is talked about nowhere near as much as Krashen, because she didn't sit down and write a one-size-fits-all approach like Krashen did.

But Krashen's one-size-fits-all approach was really just a teacher's take on Noam Chomsky's ideas.

Chomsky claimed that grammar doesn't mean anything and came up with invented lines line "colorless green ideas sleep furiously" which was not only meaningless, but was only "grammatical" in the sense that it fitted Chomsky's invented "generative grammars", a point that was countered within about a year of its publication by Lucien Tesnière with his "valency grammars" that stated such things were absolutely ungrammatical.

(I learned generative grammars three times, and it was only on the third time that I was introduced to valency grammars!)

Krashen's theories are a direct consequence of his belief in Chomsky's idea of a "language acquisition device" and his models of grammar being insufficient to help learners produce sensible utterances.

In fact, I moaned about Chomsky's generative grammars to a Polish linguist, and how the problem was put by his use of the subject - complement split of classical Latin, rather than splitting on the verb and saying "what does this verb need?", because that was the whole argument of Tesnière, who said that valency dealt with things like active/passive equivalences dead easy. What I never really understood was how they even taught generative grammars in a language which will completely drop the subject as often as not, but she'd studied in Poland and they were using Chomsky there. I don't think she'd ever heard of Tesnière!

I don't know what this is about. There was no 'attempted attack', only a reply. It didn't support your point it was a clarification of Krashen's approach to the notion of input, which you didn't use properly. It was therefore unfair to represent it as such.

Sorry -- I guess that's just me feeling vulnerable cos thinking's harder than it was last year, and difficulties in processing tend to be incorrectly perceived as external threat. Hopefully won't be a permanent thing.

On the other hand, it might just be because your idea of fun was playing Draconus or Saboteur while I had a real computer...

[Le Baron]

Yes, Krashen falling into line with Chomsky is very evident. During the pandemic someone had them together on a discussion video (it's surely still there on youtube) and Krashen is praising Chomsky and doing a sort of Jedi mind trick on him insisting they agree in all areas. Chomsky says he agrees, but looks bearded and tired.

For those who aren't familiar with the wider literature, Merrill Swain's response was the "Output hypothesis" -- basically saying that it's all well and good understanding, but that doesn't lead to being understood. If you develop a theory of what the language is, you need to test that theory, and if you're wrong, you'll get feedback and hopefully refine your theory.

Yes, and it is similar to my permanent complaint that input develops a content base and passive understanding, but that speaking or any kind of output is not just a generated outcome of input/understanding. Rather it is an additional, but related skill to be developed.

I'm certainly not a Krashenite, I posted the article 'Krashenburn' which was a scathing attack on his views as applied to teaching English to Latin American Spanish speakers in California. Yet I don't think he's a charlatan either. With regard to his name being overused, it is perhaps more in the pop langage expertise arena, and I can understand why. Acquisition by input does deliver results, but is a partial solution. Lots of people contrast it (unfavourably in my view) to school classes where the student isn't half as invested in the process as someone self-learning under their own motivation. For certain lots of input/immersion is a missing factor in average language teaching/learning, but things people say when they become 'Krashenites', about grammar knowledge being useless, to never look up anything in a dictionary, to not do any exercises and never get explanation from a teacher, is just so much ideological claptrap.

[Iversen]

I have never really thought about where Krashen got his ideas from, but the idea of a brain that automatically learns languages when fed with comprehensible input using an inbuilt mechanism could have come from Chomsky. The observation that grammar doesn't preclude the production of meaningless nonsense could also come from Chomsky, but frankly I don't remember what Krashen has written about meaningless green ideas. On the other hand the idea that any brain that is subjected to grammar gets fed up and learns nothing doesn't really seem to have come from Chomsky, who liked grammar so much that he inadvertently made it impossible to understand for ordinary humans (maybe in the hope that computers could use his constructions some day in the future).

When I wrote my final dissertation at the university in 1980 Lucien Tesnière was one of the authors I quoted with glee, precisely because his claim that grammatical analysis of sentences should start with the verbs and their concrete combination possibilities seems so much more in line with how our languages actually work - and then it's also logical to introduce the actual words used at an early stage, which inevitably entails that their meanings are taken seriously.

I can't exclude that there are some very exotic languages where it is more sensible to start with the abstract pair NP VP and then put the concrete words into the empty slots at the very last moment, but I have never been been confronted with such a language - not even a computer language. For me it's simple logic to look at the behaviours of REAL words, classify them and then build your theoretical structures on the results of those classifications - and that's how I always have treated grammar. And at this point I have to say that I simply don't remember what Krashen wrote in the one and only book of his I have read - my impression is that he was so negative towards grammar that he didn't care whether the egg or the chicken came first, i.e. whether the concrete behaviours or the abstract structures came first. But since Cainntear and others seem to have read Krashen recently they may have seen something in his oeuvre which I happened to overlook.

For me comprehensible input is a splendid notion, but I don't use it as Krashen intended (for instance I'm an ardent supporter of bilingual texts for intensive study - he apparently isn't). And likewise I see transformations as a fundamental tool in grammar, but I use them in a framework that has more in common with Tesnière than with Chomsky.

[Diomedes]

First of all, I'm not a linguist in the academic sense, and I don't have any contribution on the "comprehensive input" debate.

But as a (mostly self-taught) language learner who gets exposure to some of the "internet gurus" that reveal themselves affiliated to Krashen theories, here is what I want to add.

What disturbes me most is not the praise of "comprehensive input" they make. To me, it sounds just like one more tool (ok, I already understood it's not precisely that).

What bothers me most is their passionate insistence that "grammar is useless" when learning a language, even harmful. In my own experience, it has been quite the opposite, in the right dosage.

I can understand that in the older days grammar was incorrectly sold as "the only way/resource" to learn languages, and that, around the sixties, linguists made their own "language teaching revolution" or "may 68" and declared war on the old "system" (as all other knowledge fields were also searching for their own "paradigm shifts"). And that something good came from it. Even if their flag was "communicative approach", which I know is quite a different thing from Krashen theories

But (correct me if I'm wrong), the "new approach", or "motto" ("avoid grammar like the plague!"), which "Krashenites" seem to defend, has severe limitations, so much as the "old approach" ("prioritize grammar in detriment of all the rest!"), and the balanced consumption of very different language learning tools (reading texts with new vocabulary, grammar, etc., etc.) would be much more "healthy" in a language learning "diet".

If I weren't a somewhat experienced language learner, I believe I could be perplexed by some too radical/polarized advice (pro or against grammar) and harm my studies trusting on their "authority", and this would not be nice.

Sorry if I disturb the more qualified linguistic debate. But perhaps my thoughts can say something about how a person without academic background on linguistics (as most of their audience are) reacts to some videos about language learning made by "Krashenites".

[Irena]

But as a (mostly self-taught) language learner who gets exposure to some of the "internet gurus" that reveal themselves affiliated to Krashen theories, here is what I want to add.

What disturbes me most is not the praise of "comprehensive input" they make. To me, it sounds just like one more tool (ok, I already understood it's not precisely that).

What bothers me most is their passionate insistence that "grammar is useless" when learning a language, even harmful. In my own experience, it has been quite the opposite, in the right dosage.

Yes. And more than that, these Internet polyglots typically have a couple of strong languages and a long list of much weaker ones. When they talk about how they learned their stronger languages, it almost always involves a fair amount of traditional language study (including grammar). That's not to say that's the only thing they did to succeed (no, a grammar book alone will not make you fluent), but it's one of the things they did to succeed. And then you have others who will say stuff like "I study grammar because that's just the kind of person I am, but don't worry, you don't need grammar at all if you don't like it." Eh, right.

[Cainntear]

I have never really thought about where Krashen got his ideas from, but the idea of a brain that automatically learns languages when fed with comprehensible input using an inbuilt mechanism could have come from Chomsky. The observation that grammar doesn't preclude the production of meaningless nonsense could also come from Chomsky, but frankly I don't remember what Krashen has written about meaningless green ideas. On the other hand the idea that any brain that is subjected to grammar gets fed up and learns nothing doesn't really seem to have come from Chomsky, who liked grammar so much that he inadvertently made it impossible to understand for ordinary humans (maybe in the hope that computers could use his constructions some day in the future).

Well, I guess I'm saying that Krashen was one of those teachers who had come across some ideas from an academic and kind of read into them what he wanted to. Chomsky divorced grammar from meaning, although he didn't strictly say grammar itself was valueless. He was also talking about a "language acquisition device" pretty much when the "critical period hypothesis" was the latest big thing. I would say it's a very small leap for someone who's not a deep thinker to simplify the two things into a "learning-acquisition hypothesis" and disregarding the importance of grammar.

It's pretty common in all fields of human thought for a teacher's misinterpretation to become more popular than the sources he or she has misinterpreted, and it's a constant source of annoyance for academics to have to say "but that's not what I mean" and to hear someone who has clearly misunderstood say "no, no, I understand. What you're saying is..." and then repeat the very same misunderstanding the person was trying to address.

Talking of teachers repeating things that academics didn't say: "We all know kids learn language language best when they're young, so the earlier started, the better." No, kids who emigrate to a new country pick up the language from their peers -- primary school kids are pretty awful at learning new languages because their peers make the same mistakes as them, so the result is a weird interlanguage that is actually worse than if they'd started at high school.

When I wrote my final dissertation at the university in 1980 Lucien Tesnière was one of the authors I quoted with glee, precisely because his claim that grammatical analysis of sentences should start with the verbs and their concrete combination possibilities seems so much more in line with how our languages actually work - and then it's also logical to introduce the actual words used at an early stage, which inevitably entails that their meanings are taken seriously.

I was coming from a different angle. In computer science 20 years later, I'd been taught two types of trees: one where true data objects could only exist at "terminal" or "leaf nodes" (a leaf on a tree has no branching after all) and where the data at a "branching node" could only be metadata; and another where you could stick data on any node.

To me it made intrinsic sense that Chomsky's generative grammars made the mistake of using the first, because it meant you had a structure like

Code: Select all

      S
     / \
   NP   VP

(For those not familiar, S stands for "Sentence", NP for "Noun Phrase" (meaning here the grammatical subject) and VP for "Verb Phrase" (meaning the verb, a possible direct object in the form of another NP and a potentially endless list of adverbials)

I intrinsically understood that the reason he thought grammar was divorced form meaning was because he'd written it that way and that if he had just included the words on branching nodes, it would have been bloody obvious that the verb was the most crucial factor in determining the tree structure. This was obvious to me and I did it *twice* (in a natural language processing course and on a computer programming language module) and it was only years later that I read about Tesnière's valency grammars which confirmed *exactly what I'd been saying*.

(Now you've set me off...)

In faaaact... Chomsky's dominance despite Tesnière's attempt to correct his errors is still having repercussions today, because computer programming languages were first invented during the height of Chomskyan grammar fads and the fact that Chomsky was so wrong about language is why computer languages are so bloody hard for humans to use.

But since Cainntear and others seem to have read Krashen recently they may have seen something in his oeuvre which I happened to overlook.

Caveat emptor: knock on the head.

For me comprehensible input is a splendid notion, but I don't use it as Krashen intended (for instance I'm an ardent supporter of bilingual texts for intensive study - he apparently isn't).

But I'll say it more clearly this time:
If you say "comprehensible input", people will Google comprehensible input and they will find Krashen.
If you say "comprehensible input" therefore, you will not be understood.
It is a term that invites misinterpretation.
Therefore it is not a useful term.

[Cainntear]

First of all, [1]I'm not a linguist in the academic sense, and [2]I don't have any contribution on the "comprehensive input" debate.

[1] No, you're not; but [2] yes you do.

What disturbes me most is not the praise of "comprehensive input" they make. To me, it sounds just like one more tool (ok, I already understood it's not precisely that).

What bothers me most is their passionate insistence that "grammar is useless" when learning a language, even harmful. In my own experience, it has been quite the opposite, in the right dosage.

Yes, but as I say, "comprehensible input" is so tied to "grammar is useless" -- particularly now, in the era of Google -- that if you try to use it as a generic term, you're always going to (whether intentionally or not) drive people back to Krashen.

But (correct me if I'm wrong), the "new approach", or "motto" ("avoid grammar like the plague!"), which "Krashenites" seem to defend, has severe limitations, so much as the "old approach" ("prioritize grammar in detriment of all the rest!"), and the balanced consumption of very different language learning tools (reading texts with new vocabulary, grammar, etc., etc.) would be much more "healthy" in a language learning "diet".

[My emphasis.]

Exactly. The problem is, people are more interested in reading about "revolutionary" ideas than they are incremental improvements. If someone tries to improve the revolutionary ideas, they are neglected and (perhaps counterintuitively) they're minor improvement becomes a step on the path to enlightenment... in that literally, people start with what we do now and look for a truth in going back to before we learned everything we know now.

The market therefore militates *against* balance, despite it being provably necessary.

Sorry if I disturb the more qualified linguistic debate. But perhaps my thoughts can say something about how a person without academic background on linguistics (as most of their audience are) reacts to some videos about language learning made by "Krashenites".

No need to apologise. This is not a forum dominated by academic debates. I talk about academic stuff in order to try to push the conversation forward -- it's not an attempt to dominate the debate or to silence others. My point is to try to inform, and the proof I have been successful is people responding, even if they disagree with me (which you don't seem to do!)

[tastyonions]

Talking of teachers repeating things that academics didn't say: "We all know kids learn language language best when they're young, so the earlier started, the better." No, kids who emigrate to a new country pick up the language from their peers -- primary school kids are pretty awful at learning new languages because their peers make the same mistakes as them, so the result is a weird interlanguage that is actually worse than if they'd started at high school.

Are people who say that talking about results in the short or long term? Most who immigrate at high school age won’t end up with fully native accent or grammar.

[sfuqua]

What a wonderful thread!
It is frustrating that many of the Internet polyglots do not even seem to recognize that their is an academic field of Second Language Learning, and that it is alive and there is good work being done in it. Much of it is not very useful when one is trying to decide how to study tomorrow, but to ignore it and to keep quoting old theories from the 60s and 70s of the last century, is silly.

And Chomsky, as an undergraduate Biology, I knew enough about evolution to question whether the whole language acquisition device was a "thing" that could develop from evolution. Now, with the recent developments in large language AI models, well, I think Chomsky just goes up in the puff of smoke where he always should have been.

And I would like to say something unkind now about why many people prefer to believe the whole Krashenite program. It is because it is intellectually lazy. It removes the need for students (and especially teachers or Internet experts) to think about the learning process, since it is automatic. I am not saying that comprehensible input is unimportant, I am just saying that it is not sufficient.

Back in the 80s in the Philippines, when I worked in the refugee program for Vietnamese folks on their way to the US, I would hear teachers saying to each other at the end of the week, Pagod ako, Mag SLA lang ako ngayon. = I'm tired. I'm just going to do Second Language Acquistion today.

By this they mearnt that they were just going to walk around and do things and point at things and describe things in English. They would also do do some informal TPR. Many of these folks were briliant, interesting teachers, but the feeling was that providing comprehensible input was easy and lazy compared to actually teaching particular content in the language. And any method that removes the responsibility from learner and teacher to actually have concrete goals for their learning, is bound to be popular, because, well, who wants to think?

Itś great to see you here Cainntear. Please correct me if I am getting lost here

[Cainntear]

Talking of teachers repeating things that academics didn't say: "We all know kids learn language language best when they're young, so the earlier started, the better." No, kids who emigrate to a new country pick up the language from their peers -- primary school kids are pretty awful at learning new languages because their peers make the same mistakes as them, so the result is a weird interlanguage that is actually worse than if they'd started at high school.

Are people who say that talking about results in the short or long term? Most who immigrate at high school age won’t end up with fully native accent or grammar.

Sorry about that -- I didn't express myself clearly enough (stupid brain). The point I inadvertently skipped over (stupid brain) was the switch from talking about immigrants to talking about non-immigrants.

Immigrant children learning from their peers will be near perfect if they move into schooling the year they immigrate
The results fade off the later they arrive in primary schooling, and there's a steep decline when they're over 10. This gives us the graph that all "critical period hypothesis" stuff points back at.

So yes, primary school kids learn language very well in true immersion settings where their peers are native speakers.

(Now, for the missing change of topic!)

This does not mean that younger is better in all settings.
Primary children learn language from their peers, and if their peers have the language (as in the case of an immigrant entering school) they will do well, and in that case, younger is better. However, if their peers do not have the language then the curve is the other way round. I think I heard about a study in England that found that kids who had 2 years of foreign language at primary were overtaken by their peers at highschool who started learning in high school, often within one to two years. Primary kids who are not in true immersion are unlikely to master features their native language doesn't have (e.g. grammatical gender of non-human nouns in the Romance languages) and seem to be more likely to retain such errors later.