As some of us know Dr. Robert Patrick teaches a graduate course in education at the University of Georgia (LLED 7045 Comprehensible Input and Second Language Acquisition). He recently forwarded to me an interesting exchange with a student:
Ben,
I am copying below, with her permission but without her name, an email dialogue with one of my grad students in my CI and SLA course. I think it offers at the very least a wide window into the experiences of second language learners and how they square with CI theory and practice. The student, like more than half in the class of 32, is a native Chinese speaker, and is working on a graduate degree in English.
Student wrote:
Dear Dr. Patrick,
I learned about corrective feedback from another class this week, and there are many ideas seem to be contradictory to what we have learned from CI theory and the teaching method which followed the theory. And in the synthesis for the week, my instructor for the course gave the following idea:
“During the early, more controlled stage of the lesson, I think it is very important to recast correct answers and to request re-formulations, provide explicit corrections if recasts do not work, and draw learner attentions to the explanation, because the focus is on form at that time. During the second half of the lesson, however, when students are practicing the new form discursively, to communicate meaning, I typically write down their answers on a post-it and then go over the problematic issues with the class after the conclusion of the activity. As a proponent of socio-cultural and socio-cognitive approaches to language learning, I am a firm believer that students need to have explicit understanding of the rule which in turns allows them to control their speech (written or oral) and gradually gain automaticity. Sometimes students gain a deeper understanding of the rule when they digress from it, but the only way for them to find out how they violated the rule is by receiving corrective feedback (or “negative evidence” which refers to “what is not possible in the language”).”
What’s your opinion about it? Krashen believes that conscious knowing of rules acts as a Monitor. Should the Monitor set before or after acquiring the structure?
My response:
I take a different view, as does Krashen. I do not oppose recasting a response and using a students’ response to reformulate the question again using the correct form, but I disagree that the focus is on form in the early periods. The focus is always on the message. If we are delivering understandable messages, students will, over time, begin to produce, and their production will become more accurate in terms of form. A student will not produce structures correctly until they are ready, and we cannot do anything to rush that process.
Some our readings in Krashen this semester have addressed these issues. Dr. P
The student wrote back today:
Dr. Patrick,
Thank you for your time. I agree with your opinion. I did a little survey with my Chinese friends for their opinion on grammar learning during their English learning experience, and most of them think that it makes little difference. Even though they know some certain rules consciously (including some really complex rules), they still make errors during speaking and writing. Other students claimed that they got good grade on grammar-focused exams because they thought they have a “sense” on the language. Ironically, when they tried to use their knowledge on rules to see if they chose the right answer, chances were high that they corrected the right answer to an incorrect one.
Bob speaking again:
I think she does an incredible job of asking questions into the nature of language acquisition, of reflecting on her and her friends’ experiences and putting all of that into perspective.
Bob
7 thoughts on “An Exchange”
The evidence is quite clear that recasts can only work in one way: when modeled by teacher.
There’s a citation in the Netten & Germain 2012 paper on applying neurolinguistic principles to SL teaching that refers to an entire paper what studied recasts. When I went and read that paper, lo and behold, what did they find?
They found that the most effective way to get a kid to fix an error was with teacher-modeled recast (i.e. teacher says it right, kid repeats). Asking the student “what is wrong with your recent utterance?” is uselesss, as is “Ok here’s the rule, now apply it properly to fix your recent utterance.”
Most interesting: they did NOT, repeat NOT show that recasts do anything for acquisition. All they showed was that, basically, rule awareness and conscious focus on rules did not generally work for error remedies.
This is one of the infuriating things about reading academic papers about real-world language acquisition: much of the research is garbage, and people cite bad/limited/irrelevant research to support their views. The French teacher who works next to my room has a student teacher, and I have of course been geeking out about the stuff they give her in her Uni methods class about second-language acquisition. Most of it is flat-out wrong. When a writer makes a contention such as “explicit grammar instruction and practice are an essential although insufficient element of second-language practice,” you then want to see the research.
So you find that research…and guess what? The “study” that tests the research does the following: they give students a bunch of sample sentences (not connected in any way except via use of one grammatical device, eg the subjunctive in French ) in TL and then ask them to write additional sentences using the subjunctive. These people in other words had to inductively figure otu the rule on their own. Another group got the same sentences plus explicit instruction in how to conjugate the subjunctive. They then compare the results, and, sure enough, the people who got the explicit grammar instruction did better.
Now. It’s important here to look at what they did NOT prove. They did NOT prove that
a) people could, or did, use the rule outside of the extremely limited teaching & testing environment of the study
b) the Krashenian/Raiean inductive grammar acquisition process does not work…because the people given the sentences without the grammar instruction were not given the immediate meaning that CI teachers do. In other words, we don’t know if the inductive group’s less-able “use” of the grammar rule was because of problems figuring the rule out, or because people did not understand the sentences perfectly. Which leads us to…
…the inherent flaw in a study like this: with grammar instruction comes more establishment of meaning, because the teacher is necessarily “breaking down” sentences and doing some incidental explaining. So what worked? Was it the grammar instruction, or the fact that the grammar group had more meaning explained, or a mix of both?
There is so much crappy research like this that it would make your head spin. And textbooks and methods classes are FULL of this.
Anyway. I can send citations if people want. But the upshot I would send out to Bob and his young student is– THE DEVIL IS IN THE DETAILS. If your prof/book claims it, go read the original research…you will very often be surprised to find mere opinions, or unreal situations, or bad designs, 0r irrelevant questions.
WOW! Thank you Chris! I would give this to my dept head but we are in a good neutral place, finding things to agree on (despite the fact that we disagree on the actual core of the matter–she just confuses learning and acquisition, or refuses to understand the difference, but whatever). This helps to bolster my confidence for myself and for interacting with my students.
How exciting that half of your class are native Chinese speakers, Bob. I expect that they may be among the growing number of Chinese teachers in the USA in the future, and I am very happy that they are learning about CI methods!
The methodology described by the student is exactly what I received in all my own language study in college. I “learned” a lot. I then went to China, where I acquired. The direct grammar rules and explanations I had received did not at all transfer to fluency. Focus on form is not the way to go. Focus on meaning and the form takes care of itself over time and repetitions in a meaningful context.
I think that recasting is more for the rest of the class. Not only to model correct form, but also to restate what the person said so that everyone hears it. I think we ought to recast – especially with lower levels – after most if not all student commentary for this latter reason. Slows stuff down too. And more reps.
Recasts– if they are CI and are 100% accurate– are fine and helpful, if all hear them and understand them. Any rep of CI is useful. What does NOT work is asking people to figure out what to say, or what was wrong. BTW there is also research in Krashen (Principles and Practice) that suggests that error correction only works for the conscious mind and does nothing for acquisition.
I’ll post a comment with some more info and links later
…what does NOT work is asking people to figure out what to say, or what was wrong….
Right on and all power to the people.
And great point about Krashen, worth a copy and paste:
…error correction only works for the conscious mind and does nothing for acquisition….
That is a serious charge. The scholars who attack Krashen are not so wonderful, in fact. The emporers are losing their clothes, and fast. Run!
(I have asked Chris in a private email to expand on this and to share his knowledge about error correction and other topics with us here. Excellent. Thank you Chris!)
One thing that I have found useful is the concept of the “student expert”. (I think Ben talks about it in “TPRS in a Year”.) Today I had a student who needed help with a sentence. We worked through it, and he understood that sentence 100%. Suddenly he became the expert. When I knew he had it, I asked the people in the far corner of the room if they could hear. One of them said he didn’t quite get it, so my expert repeated it loudly to help his friend. Every time I used the structure, I checked with him on what it meant and how to say it correctly. He was the expert. In another class, one of my students knew that the answer to every question I asked him was going to be “buried”. In both cases I could see the sense of confidence and assurance expand exponentially throughout the period.
As an aside, I am once again amazed by what some of my students are doing already. My school has been in session for just over six weeks, and I have first-year students answering confidently with complete sentences. (Not all by any means, but several.) When I ask them to do partner re-tells I get everything from re-tells in English to re-tells in pretty decent German. I’m not stressing out about the English re-tells because they tell me that the students thoroughly understand the story, and it is just too early for them to output German. (Duh!) But some of them are ready because we acquire at different rates; I just provide the opportunity for expression without forcing it. In one class I do have a quartet of students who need to have a jGR reality check. I’ll enter that in just a moment.