Recasting in Language Learning

If you’re a language teacher, you’re probably quite familiar with the concept of recasting, even if you don’t know the name. And if you’re a language learner, being aware of recasting can help you learn faster. So what is recasting?

Fukuya and Zhang define a recast as “implicit corrective feedback.” Another definition of “recast” given by Han Ye in a presentation at the ACTFL 2008 conference was “a native speaker’s corrective reformulation of a student’s utterance.”

It’s not very complicated in practice. Here’s a simple example:

Student: I want read.

Teacher: Oh, you want to read?

In the above example, the English teacher communicates with the student (using a question to confirm what the student had said), while at the same time making a correction (adding “to”). The teacher may or may not choose to emphasize the correction.

Here’s a slightly more subtle example:

Student: I want read.

Teacher: What do you want to read?

In this example, while you could identify a correction in the teacher’s question, the focus is more on communication and less on correcting the mistake.

Recasts don’t have to be questions, and they can be focused on pronunciation, on grammar, on vocabulary… but they always carry with them some degree of ambiguity, because recasts are not overt corrections, and some degree of repetition is a natural part of normal speech. Will the student pick up on the correction, or will the conversation just keep moving along? (Does it even matter what the student consciously notices his mistakes?)

I believe that much of my own success in acquiring Chinese has been due to (1) getting lots of practice with native speakers, and (2) being receptive to recasts.

Here’s a typical example of an exchange that might occur (in Chinese), with a string of letters representing the focal language point:

Learner: Abcde.

Native speaker: What?

Learner: Abcde.

Native speaker: Ohhh… AbcDe!

Learner: Yes, Abcde.

The native speaker’s second utterance above was a recast, but as we see in the last line of the exchange, the learner didn’t get it. Yes, the recast was almost imperceptibly different from what the learner said originally, but recasts tend to be that way (from the learner’s perspective)… especially when they involve tones. As a learner, when you become more sensitive to recasts, you’ll hear them all the time.

Think about it… some people will pay big bucks to a teacher in order to obtain explicit corrective feedback. In actuality, though, if that person is in a second language environment, he is probably getting corrective feedback all the time in the form of recasts and not even knowing it. Recasts are great because they don’t impede the flow of information and they’re usually not an embarrassing form of correction. They’re also great because you don’t get them if you don’t get out there and talk to native speakers. They’re a positive side effect of speaking practice. As a learner, recasts are your friend.


At ACTFL 2008, Han Ye of the University of Florida presented the findings of an experiment on tonal recasting. The experiment sought to compare the effect of recasts on Chinese heritage learners with the effect of recasts on non-heritage learners. The recasts were all for tone-related errors.

Interestingly, the study found that the uptake rate for non-heritage learners was 51%, but only 28% for heritage learners.

I found this interesting for a number of reasons. The Chinese heritage learners were likely much more confident in their ability to communicate, and probably less self-conscious about their Chinese. The non-heritage learners are more receptive to feedback, but do they communicate as well?

It is likely that the role of recasts is most important in the early stages of learning a language. Our own parents used recasting on us plenty when we were children still learning our mother tongues, but eventually, either they stop doing it or we stop paying attention.

There are a lot of factors at play here, not the least of which are individual learning styles and learner personality. Recasting research continues.

I’m just one of those people that likes to pay attention to recasts.

22 Comments to “Recasting in Language Learning

  1. Recasting is the primary way I was trained to do oral error correction in the communicative method.

    My own Chinese teachers tended to use the “shout over” tone correction technique, which made me resent them as people.

    My French teachers used a similar technique, although less shouty. My French friends, in contrast, tended to take me aside and correct my errors more subtly, which was nicer, but made conversation annoying because I made a lot of errors.

  2. light487 says:

    I have to admit that I honestly get a little defensive when someone explicitly corrects me in a way that leads me to believe I have caused the listener some form of trouble or bother. This usually takes the form of “No! That’s not right!” or “You don’t say it like that..” etc. It doesn’t take me long to get over this feeling because I know the language partner I am conversing with has my learning in mind when they correct me but I think that it possibly renders me slightly less receptive to their correction that follows than if they had used a ‘recast’ instead.

    I think also there is a point, as is alluded to in the latter half of your post, at which recasts come less frequently as the speaker learns to articulate themselves more clearly in the first instance of saying something. If the mistake is a common mistake that a high percentage of people make, including heritage speakers, I would think that a listener would be more likely to overlook this and allow the conversation to continue without interuption.

    Also it would depend a lot on the social context involved. At my work, I speak with a lot of migrants over the phone who have poorly developed English skills. I would certainly clarify what it is they are asking me (recasting out of necessity) but I certainly wouldn’t make a point of it because they are customers and they didn’t call me for an English lesson, they called me to find out about their banking. So in this and other similar contexts, at least in a western country, I would predict the amount and degree of corrections to be greatly diminished.

    I certainly agree that recasting appears to be the more natural and less confrontational approach to recasting but as you say in your post, it can easily be overlooked especially if the mistake is a long standing one that has been ingrained over many years of making the mistake. Sometimes a clear and direct correction, in the right context, is required in order for the learner to realise they are making the mistake..

    Is it really necessary though? I don’t know.. I guess at the early stages, especially with a tonal language like Chinese, it is important to be precise and have a good foundation to serve as a springboard to more advanced dialgoues. However once you’re at a level of proficiency that would considered by the general population of native speakers to be adequate or competent, I’m unsure whether it is something that every learner is going to care about, unless they really are focused on bettering their language skills.

  3. I think you’re right, light487, that recasts in daily life would come across as insulting or annoying. That’s another reason why, like John said, your parents stop doing it to you when you reach a certain age.

    Also, my hypothesis would be less that heritage learners have ingrained habits that are resistant to recasts than that their goals in studying Chinese are different than non-heritage learners. Maybe many heritage learners have a “good enough” mentality, that if the teacher or language partner can understand their words well enough to correct them, then the words are “good enough” for communication and don’t have to be corrected. Meanwhile, non-native speakers start out from zero and aim not just for “good enough” but to be accepted as fluent speakers of the language, which drives them towards perfection and opens them to accept recasts. That’s just a hypothesis, though. It would be nice to hear thoughts from actual heritage students, of which I am not one.

    As a physics teacher whose students include many kids from ESL homes, I have to use recasts sometimes to correct vocabulary and usage errors for my students. I think it helps them, and it frustrates me to no end when, even worse than John’s example above, I get this scenario:

    A: abc. B: So you mean, aBc (or) Not quite, aBc. A: Oh, OK.

    I’m purposefully rephrasing the student’s awkward sentence, and the student fails to register this as a chance to correct himself.

  4. Wilson says:

    Recasting is in daily life. Monkey see, monkey do – monkey hear, monkey say. Even in conversations on the daily, repeating a question helps reinforce your answer, or if in a social multi-person environment, or a Q&A session (audience/speaker), repeating a question connects everyone listening frequently, especially good for day dreamers. Recasts are definitely everyone’s friend as you state.

    Just the other night, had a massive mis-communication with a colleague about a carbon fiber handlebar. We were both talking about the same thing, but different things and thus different understandings, though there were plenty of confirmations and yes’s and no’s and assumed communication. It got to the end, and both of us were left scratching our heads, speaking face to face in native language. Now who’s a monkey?

  5. George says:

    Being such a new student of Chinese, I have to say that recasts are far far more helpful than direct correction, although I tend to get the latter far more than the former, and for good reason: I’m barely able to get my thoughts across intelligibly and my laoshi know it.

    Direct correction disrupts the flow of the conversation and makes me pretty self conscious. After a direct correction I’m also left with the awkward judgment, “Should I repeat what I just said correctly or move on?” On the one hand, repeating it correctly helps me at the cost of being extraneous for my conversation partner, but if my conversation partner directly corrected me perhaps they don’t mind the wasted time (?).

    I think organic recasting is most likely to occur (and also more likely to be an effective learning tool) if you’re interacting with native speakers of the language in an informal context. I wonder where an Ohio resident can get that sort of environment? :)

  6. David says:

    “Our own parents used recasting on us plenty when we were children still learning our mother tongues” reminded me of my studies (not my focus but still had to visit it, learning and developmental psychology:). There is a stage, just after babies learn to say their first words, the they tend to make many grammatical mistakes (can’t really remember any examples) and many parents tend to correct the babies….to no effect. Studies showed that babies didn’t get it right the next times round even if it was corrected by the parents. Babies did get it right a year later or so…interestingly there was then in general no difference between babies whose parents who had corrected them and ones that hadn’t. This doesn’t go against that model of recasting as such (only for babies at that age). In fact in another course that wasn’t my focus:), it was drilled into us that in order to achieve behavioral change, feedback needs to be clear, specific (stop drinking coffee before 9 o’clock versus be healthy) and prompt (no use when the feedback starts: “So, five years ago you first looked left and then right before crossing the road”). Recasts on the whole are all of that…I was wondering before about whether people always have the capacities to pick it up. I sometimes get it talking to my Chinese parents-in-law. I am trying to tell the a story and occasionally there will be some recasts coming from them. My problem then usually is that I am so busy trying to work out the next things I want to say, I can’t completely pick up the recast…not enough capacity. Well, I guess that just means one thing….improve my Chinese and then I can learn more from the recasts and improve even faster:)

  7. jun says:

    “Interestingly, the study found that the uptake rate for non-heritage learners was 51%, but only 28% for heritage learners.

    I found this interesting for a number of reasons. The Chinese heritage learners were likely much more confident in their ability to communicate, and probably less self-conscious about their Chinese.”

    AH! So here it is finally expressed into words, that arrogance that “heritage learners” of Chinese which makes them believe their Chinese is great while they’re only another banana who will never get to the knees of the great non-heritage learners (like me) who surpass them in every possible way!!

  8. John says:

    jun,

    Huh? Is that a joke?

  9. John says:

    David,

    You bring up a good point. There are also orders of acquisition that have been identified for various grammatical structures in different languages. There seems to be a certain ordered process in the way the brain acquires language. But continued exposure to the target language is a necessity, and while the recasts may not have immediate effect, it’s impossible to say whether or not something is happening “in the background” of the mind, preparing for the next stage of acquisition.

  10. David says:

    Hey John

    My knowledge about that area is, admittedly (like it is with so many other areas:), very limited. Another thing that I always found quite interesting is that the baby brain within the first few weeks after birth can still distinguish between all the sounds in all the languages….but only a little bit later (can’t remember the exact time frame) the babies can only really distinguish between sounds of the language in their respective environments. That would probably explain why certain people will just never even hear the difference between some of the Chinese q-j b-p etc. It also works the other way round…most Chinese I have tried to teach a little bit of German have huge problems distinguishing between the h and the ch (as written in German). So definitely, exposure to the language is critical. About the point that the recasts might not have an obvious effect but are actually triggering something in the background I partly agree. It is impossible to always know what is going on in the background (brain scanning still has a lot of ground to cover:). I do though remember that in that particular study they did control for improvements later on in life. The study had two groups of babies. The division was made by instructing the parents to obsessively correct the grammar of the children or not . Then they checked how well the babies spoke at later times (follow-ups…that way they could see how each baby improved and how the two groups compared to eachother a few weeks later and a few years later). To make a long story short, the groups didn’t differ. If the recasts in that age period had triggered something in the background, the “were corrected-baby-group” should have been systematically better at some later stage, but it wasn’t. That of course doesn’t change the fact, that recasts are a vital for acquiring a language. The experiment is more meant to improve the understanding of how babies acquire languages. And although the findings are interesting, we have to admit that we haven’t found out that much…we only know that for a few weeks in the babies life, painful and unnatural correction of their grammar doesn’t improve their language skills :)

    Cheers

    David

  11. Sonagi says:

    “I certainly agree that recasting appears to be the more natural and less confrontational approach to recasting but as you say in your post, it can easily be overlooked especially if the mistake is a long standing one that has been ingrained over many years of making the mistake. Sometimes a clear and direct correction, in the right context, is required in order for the learner to realise they are making the mistake..”

    I agree. I think recasting, along with guessing in context, can be overrated. As an example, I was taking a low intermediate level Korean class after having lived in Korea a few years. While giving feedback on an oral exam, the teacher pointed out mistakes in the pronunciation of a Korean consonant best described as an unvoiced /j/, for which I was substituting /ch/ in some common Korean words. I had mastered the phoneme by that point and pronouncing it correctly in new words but still retained fossilized pronunciation of words learned early. After being made aware of my incorrect pronunciation, I consciously strived to correct my fossilized speech. She also noted my singsong English-influenced intonation; at home I read aloud to practice correct Korean intonation. I might be put off by strangers giving me direct corrective feedback but expect and desire it from a teacher.

  12. Rui says:

    Hi David, You are right, a number of studies have suggested that newly born babies are capable of distinguishing contrasting sounds that are non-contrastive in their language environment until four to six months later they start to lose this ability. At the age of around 10 months old, English babies’ perception becomes more like an adult English speaker—categorical, but at the cost of losing their “superpower” in discriminating /b/ and /bh/(contrastive in Hindi, unaspirated and aspirated).

    If we look at the issue from the brain perspective, current ERP (event-related potentials) studies suggested that L2 learners use the same mechanism as native speakers (Edith Kaan). But depend on the L2 proficiency and type of tasks tested (e.g. syntax or semantics), their mechanism may be delayed or to a different extent. Generally, the more proficient you are in L2 (closer to native), the more likely that your brain mechanism is similar to native speakers’. Studies using ERP also shows that discrepancy between behavior and brain response even at early stage of language learning. In other words, when the behavior test is not sensitive to reflect any change on L2 learners, their brain waves can show some slight differences.

  13. jen says:

    Of course, the downside of friendly recasting in Chinese is that the recasts could be in non-standard Mandarin. I have one friend, a Starbucks barrista, who constantly corrects me with a really heavy Hunan (Fulan) accent. It took me a while to realize this, so when I first started learning spoken Mandain, I was confused when others would recast when I would talk about needing to buy “niu lai” for my kid or wanting to throw something out in a “najitong.”

  14. Paul says:

    Both forms, direct feedback and recast are valuable and necessary in language learning. In fact, direct corrective feedback probably has a more lasting learning effect (in psychology, it is “vividness” of the learning experience that re-inforces it). Learners who resent this form of correction are just letting their egos get in the way of progress.

  15. It’s really helpful to hear these comments. I’ve been wondering about recasting a lot lately and whether I am doing it too much or not enough. I’m trying to get my students to use the skills I know they already have when they speak and write. One of them is starting to do this, the other isn’t. I’m not sure if active correction (which may make a more forcible impression but allows them to be more passive) is the way to go, or if I should emphasize a sort of “recasting-plus” approach — by which I mean starting the recast but forcing the student to finish it — or what.

    Any thoughts?

    One other totally off-topic question: I would like to get a Latin program going at the middle school level in Beijing. So far as I know, no such program exists, though I know the Center for Hellenic Studies at PKU offers Latin (and Greek), as does the Philosophy Department at Beijing Normal University. But I’m having a really hard time getting in touch with the people involved in these programs, and in any case I don’t know whether they could help me as I don’t particularly want to teach at the college level. Does anyone have any suggestions for me?

  16. wavegirl says:

    flyingfish, There are many good curricula available to purchase online for middle school Latin students. You can find out more about these resources by googling classical education websites. Some have a religious bent. Is this what you were asking? I have used Latin for Children published by Classical Academic Press in conjunction with Minimus (comic strip format) and complemented by historical novels such as The Roman Mysteries by author Caroline Lawrence. My kids have had a ball.

  17. Kelly says:

    “Of course, the downside of friendly recasting in Chinese is that the recasts could be in non-standard Mandarin. “

    Oh, I have so many Sichuan-hua and Qinghai-hua issues with my Mandarin these days precisely because I am very receptive to recasting!! I am trying to be more selectively receptive, as I keep getting laughed at when I try my Chinese out on the more educated and standardised speakers I meet in Australia!

    Seriously, though, I think you need both recasting and explicit correction. My favourite Chinese teacher always maintained the flow of our conversation by subtly recasting. At an appropriate break in the conversation, if we were in the classroom environment and if I hadn’t picked up her recasting, she would review a few key recurring errors and make me practice those. I think that takes great skill as a teacher to make those calls. I had another Chinese teacher who rarely corrected me and only used recasting, mostly because she was more interested in having the conversation with me rather than teaching me. I appreciated the teacher who could do both at once :-)

  18. Nicki says:

    I try to use recasting quite a bit with my upper level ESL students, and it works well most of the time, but sometimes I get someone who not only doesn’t “catch” the recast, but strongly resists it! For example, a real conversation I had the other day:

    Me: Is the kindergarten on break for Spring Festival now? Her: Yes. Me: Oh, I saw some kids going in this morning…. Her: Oh, some parents need to work still so the kids come for playing. Me: Oh, for daycare? Her: NO! For playing, no classes! Me: In English we call that ‘daycare.’ Her: Oh, ‘daycare.’

  19. While I feel like this entry was probably commented-to-death ages ago, I thought it was pretty exciting to see the name Han Ye up on the blog; she was my Chinese teacher at UF! She was great. I think she wasn’t doing research on my classes, though (Summer 2007), or else she was very sneaky about it. Although, as you pointed out, recasting is inherently rather sneaky…

  20. mike says:

    Is that an official term?

    When I go the convenience store (Shanghai) and buy 5 items, then ask for a bag 袋子, its always a 5 minute conversation. It’s a developing country, after all….

    Eg.

    Mike:Oh yeah, can I get a bag for that? Fuwuyuan:You want a bag? M:Yea, a bag. F:You sure? M:Yea. F:It’s 2 cents. M:Sure. … … F:Are you sure you don’t want a bag? M:….. (breathes deeply)… serenity now…serenity now

  21. Bin says:

    Hi John, May I ask if you have Han’s paper? I can’t find full text online now, the link was no longer available. Thank you very much indeed. By the way,I am interested in your masters thesis as well.

Trackbacks/Pingbacks

  1. AnthonyTeacher.com » Teaching the Language

Leave a Reply

Sinosplice and all material found herein © 2002-2014, John Pasden. All rights reserved.
Sinosplice is happily hosted by WebFaction. Design by Dao By Design