A thought on error correction

By Evan-Amos (Own work) [Public domain], via Wikimedia Commons

People have all kinds of ideas about error correction.  Error correction is the technical term for when a teacher tells someone that’s learning that they made a mistake.  Personally, I don’t think there’s any problem with error correction.  Correcting errors is just fine. It’s correcting people that causes problems. A lot of teachers think they are correcting mistakes, but their tone of voice, the look on their face, even the amount of time they wait before offering a correction are all sending a message.  And if that message is less ‘there was a mistake’ and more ‘you made a mistake,’ then sure, the student is going to feel miserable.  They might not take risks in the classroom again.  We talk about error correction without really differentiating it from people correction.  But they are very different things.  And learning how to do one without doing the other is part of the reason why teaching can be so hard.


20 thoughts on “A thought on error correction

  1. I suggest to my trainee teachers: “The more corrective feedback you give, the more you smile…the more your learners smile”. And: “pay attention whenever someone effectively helps a learner make a correction. what for? for THANK YOUs”. 🙂

    Liked by 2 people

    • Matthew,

      I think I have my favourite idea of the week:

      “pay attention whenever someone effectively helps a learner make a correction. what for? for THANK YOUs”

      When students can thank a teacher, or better yet, thank another student for helping them learn (and identifying mistakes is a big part of learning), then yes something important is happening in the classroom. Thinking of it now, my students often ask each other for help and often thank each other for their help. But I can’t think of one instance where they have thanked each other when being corrected verbally. Wow. Something to spend some serious time thinking about.


      Liked by 2 people

  2. Teachers are sometimes evaluated on whether or not they give corrective feedback, but faculty evaluation protocols don’t always consider what is happening at the student’s end. There have been a large number of studies done on corrective feedback and much of this research tends to support the view that students either do not notice error correction, or that they react negatively to it. Of course, how you do this, and why you do it, depends on the activity. Students often press me to tell them whether something is “correct” or not – and it’s not always a very easy call. When I show uncertainty, they become anxious about my skills. It takes a long time for some of them to understand that I am not looking for errors but looking for meaning, well expressed and elegantly stated. They need to trust me to do this. I believe this matter is much larger than most initially suspect.

    Liked by 1 person

    • Mark,

      I agree, that when we talk about errors and correction, we need to keep the student’s learning process in mind. Julian Edge’s “Mistakes and Corrections,” (recommended by Matthew and available here: http://www.amazon.com/Mistakes-Correction-Longman-Language-Teaching/dp/0582746264) does a good job of pointing out that when a student tries to communicate and uses an incorrect form because they haven’t been taught or exposed to the language they need in that situation, it is probably wrong to call it an error. And just trying to suss out what they are actually trying to say, as opposed to jumping in and ‘fixing the mistake,’ is enough to keep any teacher busy.

      As far as the negative reaction of learners to error correction, certainly this is the case sometimes. But I also think Penny Ur has a good point that many students do want explicit error correction. And they also want time to correct what they have produced and to think about those errors. For more from Penny, check out this link of a summary from one of her 2012 presentation’s


      In the end, I think, like you said, it’s a large and complex issue and it does often come down to a sense of trust (and mutual understanding) between the learner and the teacher.

      Thanks for much to think about,



      • Hi Kevin,

        Thanks for the reference to Edge and Ur. Both come from within established teaching traditions and are highly respected authorities. I agree with Edge, that it’s very important to try to understand what a student is trying to say before jumping in to “correct” them. I think that “correct language” is also highly contextualized – what is “correct” in a particular situation, may not be so in another. Students do not develop much rhetorical sensitivity until they have acquired substantial experience, and I think that much of this learning is not only spontaneous but defies teachability. Attempts to teach it may be counter productive.

        Ur is, of course, correct that students often want error correction. This does not, however, mean that they notice it. They may request it for many reasons, including requiring a justification for a mark on a paper – I expect this is one of the most common reasons. I would be more interested to know how much of this activity students do notice, and how they use it, espeially since error correction can occupy a great deal of teachers’ time.

        Personally, before investing a great deal of time and effort into this, I would need a more substantial argument than “students want it”. Well, I’m sure this has been thoroughly researched already. Are you familiar with Truscott’s work on this question? This was in the context of L1 English, but also relevant to ESL. Truscott argues that the claimed benefits error correction have no basis in research and that such correction is probably even harmful. This raised a sustained dialogue in the academic press in the 90s. Barbara Knoll was one of the primary scholars who rebutted Truscott’s claims, but she was unable to produce any argument other than that teachers do it and believe it is useful, and no research to substantiate her claims of effectiveness.


        • Mark,

          Thank you for taking the time to give this kind of reply. I’m (eternally) taking my dipTESOL and currently studying for the final exam part of it. So all the stuff I read and thought of at the beginning of the program is pretty fuzzy by now. This is a great chance for me to go through my notes and do some review. I think Truscott makes some good points about the difficulty of error correction (although he insists on calling it ‘grammar correction’). But his all or nothing take is a little hard for me to accept. Like Stephen Krashen’s refusal to see any middle ground between acquisition and learning, I think by challenge the idea that there is something inherently useful in error correction, Truscott helps challenge teachers to take a look at their practices and make sure they are justifiable. And just in case anyone is interested in what Truscott has to say on the subject, you can check out most of his articles here:


          Not to get too deep into a debate which has been raging for a long time, I guess I’m a bit more in the camp of R Lyster, PM Lightbown, N Spada (http://people.mcgill.ca/files/roy.lyster/LysterL_S1999_CMLR.pdf). If an activity is focusing on form and what the students are shooting for is accuracy, than I think timely and clear feedback is useful. But I have a feeling we are mostly on the same page here. It really does depend on the student, the activity, the way the class has been going up to that moment, and a million other things. I hope that when I do work with a student so that what they have to say is more understandable (or as you say, “well expressed and elegantly stated”), I am managing to keep track of as many of these issues as possible so the student does have a chance to learn from the experience.

          I’m basically talking about oral error correction here and I guess I could put down a few things I try to keep in mind when correcting errors:

          – Error correction is a no-no when students are engaged in a communicative activity where they are negotiating meaning with each other.
          – Give students a few seconds to think about what they have said before doing any kind of correction.
          – When possible give students a second and third chance to say what they said before correcting. I find that many errors are simply slips and disappear when a student give it another try.
          – If possible use what a student has said as a base to build towards something else. What I mean is, focus on what is right in the sentence and then clearly let the student know where in there is room for improvement. Techniques like loading a sentence on your fingers (one finger for each word) so they have a visual cue of where they might need to insert words or what word to change).
          – Make a conscious decision to use the moment to teach the class or to teach the individual.

          But even these ‘rules’ aren’t written in stone. Like so many things in language teaching, the more I read and think about a particular issue, the more I feel that my answer today will not necessarily be my answer tomorrow. And that it’s the process of reflection and experience which is more important than any particular theory.

          Thank you again for the chance to think things through. And I hope that if any other readers have some beliefs about error correction, or some ideas for how the go about it in their own classroom, they might share it here in the comment section as well.


          Liked by 1 person

  3. Pingback: Day 63: Its National Grammar Day! | A new day, a new thing

  4. Pingback: Day 64: 2285 | A new day, a new thing

  5. ‘Error correction’ is one of those very sweeping terms that needs to be broken down. As well as the fluency v. accuracy angle, there are the different kinds of error to consider (grammar, vocab, pronunciation, syntax, and more…) We also then need to consider if it is a ‘slip’ (or a ‘typo’ for written work), a case of trying to use unfamiliar language, an L1 influenced error, or something else.

    The main probelm area for me are spoken errors – no-one likes to be corrected when they are speaking. Written wrrors can be corrected or addressed (a better word for it, I think!) much more discreetly whereas oral correction tends to be more on the spot. Even when we note down errors and noard them a few minutes later, a students can still think “uh-oh, I said that”

    Then there is the noticing or errors in the first place. I was recently observed and asked in the feedback why I had not corrected a girl when she said “I had wok-ed up late…” I could have sworn she had said ‘woken up’ but my oberver was quite adamant that she had said the error. Before we could even discuss the appropriate way to correct the error, if at all, we had to discuss whether or not there was indeed an error at all!

    Liked by 1 person

      • Dave,

        Thanks for a reply which illustrates your point so perfectly. Indeed, a slip isn’t an error and even in a piece of writing, it seems that pointing out each and every ‘slip’, let alone forcing a learner to wrestle with each and every error would be less than 100% effective.

        On the oral side of things, I think John Fanselow makes a lot of sense when he pushes for recording and transcribing conversations, not just for reflection on the part of the teachers, but on the part of the students as well. In some ways, the errors that the students can hear and notice themselves during the transcription process are most likely the ones that can benefit the most from conscious language study. And recording gives the responsibility of their own learning back to the student. I do think some spoken errors can be addressed (like this term much more than corrected, thank you) in the language classroom. But like you said, no one likes to be interrupted when talking. And if students can never really figure out why a teacher corrects sometimes but not others, we just end up pushing the classroom world one step further away from the ‘real world.’ Because in the real world, people don’t correct each other’s grammar. Or if they do, they are avoided at parties…hey…maybe that could explain why sometimes I find myself alone in a corner.



        • Dave and Kevin,

          You raise a very interesting question about error tolerance, which following traditional institutional expectations would probably be near zero. The irrationality of this should be clear to anyone. In the wide world, everyone seems to establish their own standard of tolerance, and this is often contextual as well. Situations where we impose zero tolerance are few and far between.

          Perhaps we could allow our students some autonomy here, and allow them the same liberty we insist upon for ourselves to say when enough is enough.

          Liked by 1 person

  6. Love, love, love this distinction! Yes, this is it. You have finally put the words to what I have thought but never articulated. I was making it much more complicated in my mind. I can imagine myself bringing this to my course participants: are you correcting the error or correcting the person? Such a juicy question. Thank you for this Kevin!


    • Hi Josette,

      Thank you. I think I mentioned today that I was reading Carl Rogers’ ‘On Becoming a Person,’ and you mentioned that the book might have influenced this post. I think it probably did influence my thinking when I wrote this and I hope it continues to do so. The idea that a person grows and changes when they are most free to be themselves (I imagine this means being able to embrace and accept things like making mistakes as well), which is at the heart of Rogers’ philosophy, probably requires a very different attitude towards error correction than what I’ve practiced up until now. Of course just knowing that there is a better way to address an error, and actually finding ‘the right way’ for a given moment/context are two different things. But at the very least, focusing on the error in a nonjudgmental way might be a good first step in the right direction.


      Liked by 1 person

    • Howdy MFT (My Friend Tyson),

      Not sure if the whole not-mad-when-smiled-at thing is true or not. I’ve had some students who knew how to flash the kind of smile that was clearly anger-tinder material. On the other hand, hurt feelings? Let’s hold a double blind experiment at the next conference we show up at. We can hurl insults while smiling half the time and collect data regarding the different reactions. Although we might get run out of the building before we gather enough data.



  7. Pingback: Some thoughts on Error Correction | ricardo barros elt

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s