My response to Butzkamm article re Krashen

A while back, Paul Widergren kindly passed on to flteach members a url to reach an article by Wolfgang Butzkamm. The url is
The title is The Language Acquisition Mystique: Tried and Found Wanting

I finally got around to reading it and found it to be a perfect example of the defense of a paradigm engaged in by those holding the line, so to speak. Of great interest to me was Butzkamm sites three publications of Ron Sheen, someone who was prolific on flteach and finally got expelled for ungentlemanly behavior, and Bob Ponterio and Jean LeLoup do not easily strike members from the flteach rolls. I would not put that fact into an academic article, but considering Ron’s extremely vicious attacks on those who disagreed with him, and considering that this blog is personal, and considering that I was the target of a good many of his attacks, I think it entirely appropriate to include this tidbit of gossip. More than that, Ron’s tone was not so far from Butzkamm’s and that is what I turn to before we get to the meat of his argument.
Butzkamm starts off by setting himself up as a gatekeeper, decrying Krashen’s ……… oh, did I mention? This whole thing is an attack on Krashen, and I use the word “attack” advisedly. As a gatekeeper, he has decided that Krashen’s coining of the terms acquisition vs learning set in opposition to one another is nothing more than the more usual “intentional vs incidental learning”. What he missed was that acquisition is not incidental, it is part of a natural process (lots more on “natural” later). He makes bold claims using expressions like “unassailable fact”, “arrogant attacks”, “there can be no doubt”, “all good teachers know”, and others. He clearly wishes us to see him as an unassailable authority. This is important and I place it near the beginning of my review of the article because I perceive here a desperation, perhaps born of the turn to Comprehensible Input and Communicative models of teaching across the world. I myself have seen trps, a CI method, burgeoning from the early 90s on through the early 21st century when people were still asking what tprs is on to today with about 6000 practitioners of tprs on a tprs listserv, certainly only a fraction of those implementing the method.
So on to what Butzkamm objects to and then to what he advocates. This is not so easy because he makes statements like: “
“The idea that we first learn a new rule, and eventually, through practice, acquire it, is widespread” (Krashen 1982: 83). However, I have yet to find a methodologist of the 20th century who advises us to do so. Ever since the days of Harold Palmer, of Jespersen in Denmark or Philip Aronstein in Germany grammar rules have been presented only in close conjunction with demonstration and practice.” He does state he is speaking of Europe, but what I am unclear on is the distinction between “first learn a new rule, and eventually, through practice, acquire it” and “grammar rules have been presented only in close conjunction with demonstration and practice”. Maybe it’s just me.

Here is one point where I am unsure of myself, where Butzkamm writes, “There is no empirical basis for the assertion that the monitor, i.e. declarative knowledge does not convert into procedural knowledge.” To me, that yet remains an area that needs to prove a negative and my grasp of good research design is not the best.
Butzkamm defines acquisition for himself; it is informal, intuitional in the crib or on the street. That is far from what Krashen means but it fits perfectly well with the objections to CI over the years on flteach, the notion that “picking up” a language results in despised street talk, similar in tone to the attacks on non-standard English I’ve read so often on flteach. When I’ve suggested we as teachers learn more about English grammar, incl. the speech labeled non-standard, in order to teach our students coming from non-standard language homes, I’ve been not just disagreed with but vilified for “wanting to teach non-standard English in the classroom”. Shades of the Ebonics controversy!
After defining acquisition as learning in the crib and the street, Butzkamm defines its opposite, direct instruction, as where the teacher presents, practices, talks about, and explains the text. He offers that learners should first encounter features in the text and only then discuss and analyze and practice. This is what he labels the conventional classroom and I can only agree with him, but it certainly seems to contradict the above quote: ““The idea that we first learn a new rule, and eventually, through practice, acquire it, is widespread” (Krashen 1982: 83). However, I have yet to find a methodologist of the 20th century who advises us to do so.” He says no, we don’t do that and then says that is what direct instruction is. Am I the only one who finds that confusing?
In fact, Butzkamm is amazingly traditional, what we now call, thanks to Terry Waltz, legacy teaching. In order to learn a language, a fanciful tourist must “break down” an expression, CONSCIOUSLY or SUBCONSCIOUSLY, in order to analogize its features. Only then does real learning take off. Analogizing is required to “multiply our production”. He wants the learner to understand the structure, “must know how to put the message together.” In the word “know” is the mystery. Since Butzkamm regards language as a skill, the first law of skill learning is that we learn what we do, so speaking, aka output, is required. What Butzkamm does not explain is what output is based on or modeled on. Apparently, it is a matter of rule application or “rule-getting and rule-using”. [see my blog entry on this at]. If we do what structure drills have us do, we are right back to ALM, Contrastive Analysis, IOW, cognitive code, all methods found wanting in the past that Butzkamm so reveres.
For Butzkamm, the link between language ability and perceptual-motor skills is clear, so let’s refer to Frank Smith on this: skills are physically based and can therefore be practiced with benefit; language is a faculty of the mind and the mind is not a muscle, it works differently. Butzkamm’s whole approach to language learning is based on the notion of skills and he compares language “skills” to the skills of typing, Morse code, driving, or playing the piano. He says Krashen has not psychological theory of learning and then he offers us this: skills psychology. That’s fine, but he presents it in his usual mode: as a fiat. He describes the process of language learning in a vague way, perhaps due to limitations of space, as a matter of neurons dropping off as they are no longer needed as the feature is established or mastered: “As we attain mastery, some neuronal connections just become silent.” Whenever I see the word “just”, I become suspicious someone is asking me to assume principles as a priori. It sounds vague.
Truscott, in contrast to others who describe the non-interface theory of K as “discredited”, states that the burden of proof lies not on the non-interface proponents but on those who claim a result from a merger of conscious and unconscious learning, that implicit grammar comes from consciously acquired knowledge. This can be found on the web here:
Consciousness and Second Language Learning
John Truscott – 2014 – ‎Language Arts & Disciplines
It is important to recognize that the burden of proof here does not fall specifically on … Given this complication, the natural default position would seem to be the …

Having read a good deal of SLA theory, I was nevertheless taken aback when I googled non-interface and found numerous articles I skimmed, noting the extreme complexity of the arguments. When the field is this complex and deep, it seems premature to dismiss any substantial body of work as discredited. And that brings me to my overall observation: a great many supposedly objective researchers in SLA seem dedicated to preserving legacy teaching with its emphasis on academic learning and awareness of structures as opposed to allowing the brain to do what it clearly does in any natural situation: learn a language, not by analyzing and analogizing in a conscious way but by doing what K has dubbed acquiring. In my blog I wrote up a review of Larsen-Freeman who takes this “it stands to reason that….” approach, an appeal to “common knowledge”, at least among fl teachers.

Full version of review of Larsen-Freeman

Contrasting Long’s statement and Butzkamm’s is illustrative of the problem: B quotes L as saying a grammatical syllabus runs roughshod over “readiness to learn”, the natural order of acquisition and B counters with an airy claim that fl teachers all over the world have used the syllabus of their textbooks quite successfully. Period. Where is the vaunted empirical evidence? More telling, B labels the concern for readiness as Rousseauist, revealing his bias against any naturalistic approach. “Language acquisition theorists espouse a Rousseauist view of the language learner.” p. 90. Further, he says, “”…arrogant attacks such as Long’s against time-tested teaching practices.” But later he says, “… there can be no doubt whatsoever that explicit knowledge can speed up the acquisition process….”

And here is the key as far as I am concerned. Butzkamm strikes out, as so many have on listservs like flteach, against any way of teaching language that bases itself on a natural tendency to acquire language, not just among first language learners but among adult learners as well. Butzkamm defines language as a skill and from that flows the emphasis on conscious learning, on practice, on classroom discipline and structure, on correctness, on analysis of language….. the only item in the arsenal he doesn’t mention is memory. Just as Truscott challenges scholars like Butzkamm to show how it is that implicit aka acquired grammar comes from consciously acquired knowledge, so might we ask Butzkamm how it is that innumerable people throughout history have acquired/learned another language without conscious effort, classroom study, textbooks, or other academic paraphernalia. (Challenges to this I’ve seen cite failure to master formal written features of L2 as evidence of insufficiency – the stuff of school masters and AP examiners)
The empirical evidence Butzkamm offers for his proposition that input works for advanced students but not beginners is “didactic common sense” and “pedagogical experience over many decades”. Unbelievable. It matches Larsen-Freeman’s “it stands to reason”. Where is the empirical evidence that practice works?
What would empirical evidence of CI’s effectiveness be? How about OPIs given to students? Frequently I’ve written of my own project that started sometime in high school, in the fifties, when I began using my knowledge of several languages to casually check out people’s proficiency. You can imagine how many hundreds of people I have informally interviewed as to their language skills. This really got focused the year I was a substitute teacher and covered fl classes in several high schools. I would first gauge the students’ abilities in L2, always lower than their time in class should warrant, then I would begin teaching a little. Invariably, they were excited to find they actually could understand this language they had been studying for one, two, or three years, thus demonstrating that children have an innate desire to communicate. To cap it off, I was given a five week assignment to cover for a teacher I had subbed for before. The classes were terrible, 2 French One, 2 French Two and a Spanish One, but given five weeks, I took the challenge to turn the classes around. I accomplished that, creating a classroom environment where comprehension was expected and delivered.
The following year I took over a Latin program and the students were almost without exception excellent; they had mastered, in the martinet sense of the word, the grammar. When I began speaking to them and when we read stories and discussed them, the third year students would exclaim, “So that’s what that’s for!”, recognizing the function of a grammatical feature. This mirrors Butzkamm’s assertion that the adult learner and the L1 learner must learn to recognize who the “t” of “je t’aime” refers to (I’lll stay with English: the “you” of “I love you”), just as the tourist must learn to parse s’il vous plait so he can say l’hotel me plait, etc. The child MUST attach the change in grammar of ball~balls, i.e. -s to the change in the situation. Conscious, intellectual, analytical, academic……. oh, joy! Oy veh! Which begs the question of learning via grammar: they had studied grammar and now they were understanding spoken and written Latin, so the answer is to teach grammar. But that is not what happened. What I did was build on scraps over several days. Those who have been through a tprs presentation of just a few days such as Fluency Fast will tell you they have a sense of knowing L2, albeit limited, but L2 “falls out of their mouths” (old tprs phrase).
Here is the problem for me: I know when something is acquired and when it is not b/c it “feels” acquired, i.e. it is a subjective phenomenon. For instance, the subjunctive in Spanish in contexts such as “suggest that”, “wants X to”, etc. but it is not yet acquired following negative except “no creo que”. Some genius will figure out a way to test for the distinction between learned and acquired but until that happens, that will be a weak spot in any attempt to make K’s theories part of the canon.
The essence of B’s position, shorn of his anger toward those who would diminish the role of direct instruction in grammar, is “Ultimately those on the acquisition track, as we all know, will overtake classroom learners who must set off faster and must make more out of less. Said another way, for classroom learners, there is less primary matter to begin with [= input], which means that they must build more on constructed matter – through carefully crafted grammatical exercise. “ He says, “Let us learn from the past” and “… the overwhelming evidence of history…” Not really empirical evidence, is it?. The bracketed item, = input, is my interpolation to point out that what CI methods like tprs do is supply just the think Butzkamm admits boosts acquisition, input.
The major theoretical flaw is the equation of skill and ability, the latter being what the language faculty is, not the former, which is based in physical action. We might call the distinction between the two K’s “theory of learning”. Butzkamm’s proof is to refer to his own writings and those of Vivian Cook, but Cook writes about distinctions between decoding, i.e. understanding the message, and code breaking, i.e. figuring out the grammar, figuring out that “you” in “I love you” means the child L1 learner when mom kisses her but that “him” in “I love him” means someone else when mom says that and kisses someone else. Krashen’s insight was to build on Chomsky’s Universal Grammar and Language Acquisition Device conceit to hypothesize that we learn language like we learn to walk – it is part of our biology, i.e. instinctual, and untutored, and that LAD continues operating into adulthood, if a bit more creakily. Imagine someone claiming they taught their child to walk. At least that is my understanding of Krashen. B claims K conflates the two levels: code breaking and decoding, but fails to at least understand K is saying that code breaking is an intellectual process for linguists to engage in and not the way humans have learned other languages over the eons. And decoding as a term has been rejected even by Latin teachers! Maybe B is using it differently, but then he is guilty of what he accuses K of: ad hoc definitions.
In the end, I can summarize B is by characterizing him: “we have always set up our own criteria for success via our tests and base our work on centuries of trial and error, while the acquisition people want to throw all this over in the face of empirical evidence that their practices do not work (their claims are unsubstantiated and not supported by classroom research).” This is classic paradigm defense.

I would love to hear from anyone who differs with me on this or with anyone mentioned in this review.

Sheen, R. (2003). Focus on form – a myth in the making? In: ELT Journal 57, 225 -233.
Sheen, R. (2005). Focus on formS as a means of improving accurate oral production. In: Housen,
A. & Pierrard, M. (eds.). Investigations in Instructed Second Language Acquisition. New York:
Mouton de Gruyter, 271-310.
Sheen, R. (2007). An examination of the validity of the principles of incidental learning and developmental
sequences. In: Pawlak, M. (ed.). Exploring focus on form in language teaching. Poznan:
Faculty of Pedagogy and Fine Arts, 170 – 193.


  1. Judy Dubois says:

    Hi, Pat. Thank you for this thoughtful and insightful critique of Butzkammen. When I read of Larsen-Freeman ” she pictures a gifted teacher explaining very carefully, with lots of examples, the way indirect objects are handled in the TL and then, with lots of practice, sees the students – at least some of them – reach some level of proficiency” I couldn’t help thinking of my own experience, a story I’ve often told to explain why I adopted Comprehensible Input methods late in my teaching career.

    In my lycée I was considered the “Queen of the Passive Voice”. I had these great diagrams that my colleagues loved to copy which I used to explain the mechanism of the Passive Voice in English. Of course, if my students didn’t become proficient, it could only be because they were “weak/unmotivated/not serious” etc. Then one year I was given an excellent class of highly gifted students. They were smart, highly motivated and serious, hard-working students. I gave them my diagrams, did a great job of explaining them, they did all the exercises I gave them as homework and aced the test on Passive Voice. But …. I could not help but notice that once the test was history, they were not using the Passive Voice in their writing, and if by chance someone did hasard it, the form was incorrect. In other words, they had “learned” the structure for the test, but they had not acquired it.
    A few years later I found myself teaching a group of boys who had been failing English and were a major management problem for their teacher. They were weak/unmotivated/could care less. We began watching The Lord of the Rings in English, translating the English subtitles and discussing the situation. The entire Prologue is in the passive voice. I asked them “How many rings were given to the Elves?” etc. We circled all the information in the prologue, which naturally involved the passive voice. I did not mention the structure or give them any explicit instruction about it. They did not practice it, since their answers were usually one word phrases.
    Me: How many rings were given to the Dwarfs? Students: Seven.
    Me: Where was the master ring made? Students: in Mount Doom.

    A couple of months later, we were still watching The Lord of the Rings and one day we were discussing the Ring’s powers. I wanted to ask, “Who was the Ring made by?” but realized we had not used the passive voice for some time and thought it might be rusty. So I asked “Who made the ring?” Immediately one of the boys said, “The rign was made by Sauron.”
    You can imagine how delighted I was. I rewarded him with 5 bonus participation points. His friend wanted to know what for. I said, “Because he used the passive voice correctly.” His friend snorted, “He doesn’t even know what the passive voice is.” I replied, “I don’t care, as long as he uses it correctly.”

    That’s an anecdote, it’s not data, but it illustrates perfectly for me the difference between acquisition and learning and why Comprehensible Input methods work better than explicit instruction and practice.

    1. Pat Barrett says:

      That story goes right to the heart of how we learn languages. I have saved many such stories over the years but not catalogued them, so it means going through several thousand e-mails from past years and locating such stories – most on the moretprs listserv, fewer on flteach but more and more, and lots on LatinBestPractices. I’d do one of those vanity press books, what they now call self-publishing with Lulu or whatever. I think lots of tprs and CI teachers would enjoy reading them and they might convince the grammar royalty to try a little CI. And I, of course, have any number of such stories.
      I’m still trying to make something snarky out of Butzkamm’s name on the pattern of dashcam.

  2. Lizette Liebold says:

    I love Judy’s post. We as teachers have so much practical knowledge based on our experience in the classroom. I enjoy hearing personal anecdotes. I have one that is similar to Judy’s. For many years before my introduction to Krashen and TPRS, I was a traditional textbook teacher. When we arrived at indirect object pronouns, I dutifully introduced and practiced them. Because of the organized and clear presentation and practice of the grammar item, students did well on quizzes. However, for the rest of the year, I was always puzzled by the fact that students never used IO pronouns in speech or writing. At one point in my career, I simply stopped teaching them! Why waste the time? Then at one point, after “converting” to TPRS, and without any explicit grammar teaching, my students began using IO pronouns in both their writing and speech. Anecdotal evidence? You bet!

  3. Pat Barrett says:

    This encourages me to start gathering my own anecdotes and then setting up a daily search for other great stories. If anyone knows of any such stories, post them to my blog ( or make a comment on this entry.

Leave a Reply

Your email address will not be published. Required fields are marked *