Does it help to correct mistakes in language learning ?

Krashen feels that correcting language mistakes is a waste of time. He has done research to show that it does not lead to improvement. He is apparently also concerned that too much correction will create what he calls affective filters, in other words, emotions, that prevent us from learning. He may be right. I do not know.

I agree that language learning is a process which relies a great deal on our attitudes and emotions. I have also observed that correcting people while they speak has little effect, since the speaker is usually focused on speaking and cannot take in the correction. The learner usually repeats the mistake almost immediately, even after repeated correction. Apparently correction is similarly ineffective even in the long term according to Krashen.

I also know that people often know when they have made a mistake, and may not make that mistake all the time. Often they will gradually overcome their mistakes on their own, and come to speak more correctly. They may also be quite satisfied with their current degree of accuracy in producing the language.

Often when I post a video at Youtube, in a language other than English, there will be people who come along and correct me. I must admit that I find it annoying. I appreciate people telling me how to say things that I obviously struggled to say, like "wash the dishes" in my recent Russian video. However, telling me that I overuse this or that word, or misuse certain constructions, is essentially useless and unwelcome.

Making mistakes is a valuable part of learning, but only if we are aware that we are making mistakes. So it is helpful to be made aware of mistakes. So what should we do?

My rules are as follows.

1) Never correct unless asked by the speaker to do so.  The idea of perfect strangers correcting my use of language, on youtube or anywhere else, strikes me as just rude, and certainly not helpful.
2) Even if asked to correct, keep it to a minimum, or the learner might become discouraged.
3) Correct writing much more strictly than spoken language. People often make mistakes while speaking that they may not normally make. In other words it is hard to always perform at one's best while speaking.
4) If possible, provide a few comments after a conversation, (and that is what we do at LingQ) and not during the conversation.
5) The objective of correction is not to bring about an immediate improvement, but to help the learner notice what is happening in the language, and where his or her gaps are. This provides an incentive to notice the language more, while listening and reading.

So, in my view, some correction is a good thing if done right, and if we do not expect immediate improvement, nor consider mistakes to be bad. In fact mistakes are good, we need to be made aware of them, but in the right way.

Comments

[igordesu]

well, purely on the issue of making mistakes in your target language, I think that just means that you need more input. If you make too many mistakes or struggle with something too much, then it just means you need more input.

And making mistakes probably doesn't harm too much anyways. It's like, if you get enough input, it'll eventually override the mistakes. Example; try to speak incorrectly in English. It takes conscious effort (and is almost painful, lol), whereas speaking correctly does not.

[Colin]

Agreed. Mistake correction, while it can have it's place, is largely useless. If I notice the mistake or seek such notification out then I am more readily able to internalize the correction. Of course we all would prefer to use the language "correctly" rather than "incorrectly", but those corrections are best discovered by the learner naturally, rather than pointed out (often uncomfortably so) by an outsider.

Steve, different topic but related to Krashen... in the past you have talked about bilingual education, French immersion in Canada, and literacy. I came across this NPR roundtable, which included Stephen Krashen, about bilingual education for immigrant children in the US, in which they touched on those topics:

http://www.onenation.org/0108/082001a.htm

I would love to hear your thoughts on this. Maybe in a future podcast/video.

[Steve Kaufmann]

Colin,

I do not know if I will have to do one in the near future on this subject, nor do I have enough experience or knowledge to really talk about it. My feeling is that immigrants to a country should study in the language of the country and any effort to maintain the language of the parents should be up to the family.

Learning the main language of the country is simply too important for their future, and to their integration to be left to chance. Nor do I think that the children of immigrants need to be under any obligation to learn the language of their parents. They should be treated like any other school children. Nor should, say, Hispanic children be treated differently from children of other origins.

I think this is better for the kids and for the society as a whole.

As far as French immersion in Canada, my grandchildren are in French immersion,but I am not convinced it is a good solution. These kids do not end up fluently bi-lingual, and the program ends up being somewhat elitist. They should improve how French, and other languages, are taught in the school system, for everyone.

[Chris S]

It's completely correct how correcting mistakes gives no help and if anything hurts the learner.

I've gotten in the bad habit of correcting my wife on really small unimportant things the are non existent in Polish like different question order and articles (a.the.an) and all it does is annoy her and really since it annoyed her it never got better.

But what was really negative about it, is now when she's around other native speakers, she knows she makes some of these mistakes and now she tells me she's become completely afraid to talk in front of them. Huge mistake on my part and I hope I didn't put some kind of blockage on her improvement.

[Max]

It's not the method that makes a correction good. It's just that some people need it, some don't. In chess some chess-players bother to do the post-mortem analysis, some immiedately focus on the next game.

[Keith]

Self-correction is the best type. One can overcome bad habits by self-correcting.

But how do mistakes creep into the learner's language? I think, many times the learner knows that what they are saying isn't correct but since they aren't corrected they may become used to what they made up and then forget that they never checked it for correctness. What they said was understood and accepted.

Therefor, it is up to the learner to be responsible for his or her own language learning.

[Friedemann]

Steve,

I would never point out your mistakes in a foreign language. But in English I will. You wrote: "I must admit that find it annoying." C'mon Steve, shape up!!! (LOL!!!)

Friedemann

[Emilio Wuerges]

I think this is a question of trust.
If the student trusts the teacher, he will learn. Else he won't.
But trust is something that has to be built. It takes time and effort.
At least, this was my learning experience.

[lyzazel]

Could you please as well point out to the exact study of Krashen you are talking about?

I have always assumed that it helps to correct mistakes and even though it can sometimes be a little bit annoying (well, the time when I found it annoying was when my teacher used to correct me on almost every word, usually not when other people do it) I still think that it helps me learn by exposing me to correct ways to say things which I often remember afterwards.

[Pedro]

I think that normally the students are asked to speak too soon, when they aren't ready, ao the mistakes are more frequent.
Another important point is that sometimes the learners know the correct way to speak but they aren't used to produce it while speaking.
In any case, the problem can be solved by having more input.

When I'm speaking I don't like to be corrected. Because I'm focusing in communicating ideas and in being able to understand what has been said.

I think that the only value of error correction is when it comes to notice. And in this way I think that LingQ is very effective. Because you get reports from your activities.

You don't mind while you're producing the language. You don't have to take notes. But you can know what are your weak points.

[Steve Kaufmann]

Thanks Friedemann, I appreciate errors and typos being pointed out, since this is supposed to be about language. Nevertheless, it is difficult for me to crank all this stuff out without making typos, so my readers are my quality control department. That is a different story from pointing out usage errors in a foreign language.

[Steve Kaufmann]

Friedemann, I should point out that you will find more typos in my comments since these do not allow editing.

[Liam]

"However, telling me that I overuse this or that word, or misuse certain constructions, is essentially useless and unwelcome."

Sounds like you need to, as Chopper Read would say, harden the @#$% up. Pointing out mistakes will help you correct them next time. The goal should be to speak as perfectly as possible.

[Chris S]

The guy speaks 9 languages or something and people continue to tell him how to learn...

there are about 5 other people with a presence on the internet that PERHAPS have any room to criticize Steve's techniques and advice...

[Ilya]

Steve, you wrote:
“4) If possible, provide a few comments after a conversation, (and that is what we do at LingQ) and not during the conversation”.

Yes, the comments at LingQ are offered to the learners mostly in the typing and, in many cases, they are NOT sent to the learners soon enough after the conversations.

I am a learner of English and a tutor of Russian at LingQ, and thus I take part in the conversations, conducted via Skype, both as a learner and a “teacher”. When I am a learner of English, I would record the conversation. Soon or immediately after, still being “warm” from the effort of expressing myself, I would play the record back. And I would usually learn something useful. Using Steve’s metaphor, the recording is highly resonant in me just then.

And I think I would get more from those recordings if my tutor’s corrections were in there, that is, if they had been spelled out during the conversations. I agree that I might not take in too many corrections articulated DURING the conversation, but I would prefer to have many of them AFTER. At least I would think of a compromise.

I would only be happy if certain corrections, say the pointers to the incorrect stress or incorrect intonation, were made, and made repeatedly, by the tutor’s voice, rather than by the typing. For example, in my last conversation with Steve, I incorrectly stressed “content”, “record” and “comments". Voice corrects this better than typing, and it is better to correct on the spot. By the same token, why don’t we teach a learner, at least the one who is asking for it: “Look, you modulate it so (trying to mimic the learner’s intonation), while I would do it so”. It would make a high resonant recording that could teach me something.

On the other hand, below are two typical corrections from a report on a conversation sent to me by an English tutor at LingQ not long ago. My misuse is in the citation marks. The two corrections constitute exactly a half of the report. I believe they have taught me nothing.

1.“Many Russian live” -- Many Russians live... (Every time we use the word "many," the next noun is plural).
2. “We visit Washington” -- We visited Washington (speaking in the past tense).

Don’t get me wrong. I liked that conversation and the tutor, and have learned something from the recording, in which my speech the tutor’s corrections sounded in high resonance.

[Sam]

This is a tricky one and I think it depends on the learner's preferences and who they're around.

I personally dislike being interrupted for corrections when I'm speaking (unless I've clearly said something offensive). On the other hand, sometimes the annoyance factor helps me remember the correction.

I like the idea of being allowed to speak and write freely, and have any corrections in my speaking or writing told to me after.

I'm JUST starting to learn a tiny bit of German, but had to tell my fluent German-speaking girlfriend that her numerous caveats and fussiness over pronunciation was getting a bit intimidating and off-putting.

[Igor]

It does not depend on the learner's preferences, who in this world would prefer to crawl if he can run, also it does not takes "more" input, it takes "continuous" input, you see, you start from a point "A" and trying to get to a point "B" anywhere in between points you'll make mistakes because speaking and writing (output) does not comply with reading and listening (input), on beginners level you might need intermediate speaking skills, on intermediate level you might need advanced speaking skills that you'll acquire later, so if you know you acquire them anyway why bother at all, live with what you got, the future is bright, time is always on your side.

[Steve Kaufmann]

To me,the important points about correction are:
1)the learner should decide whether he or she wants to be corrected, when and how.The attitude of the learner towards the language, his or her enthusiasm and confidence is more important than any little error correction.
2)correcting is of relatively minor importance in the total process of language acquisition.

Liam, it is not for you to tell me, a learner, what my goal in learning a language should be. My goals in Russian are understanding what I read and hear, and communicating. I expect that my accuracy will improve gradually, with or without correction.

Ilya,

I have 3 Russian tutors, 2 Italian tutors, and 1 Portuguese tutor at LingQ, and have had discussions with 2 German tutors. I usually get my report fairly quickly. The length or usefulness of the report varies, as does the nature of the conversation I have with my tutors. I treat the report as content, from which I save words and phrases, so that I can start to notice these paterns. The report usually contains words and phrases that I know I need to improve on, and it does not matter whether I get the report within 15 minutes or 5 hours.

Most tutors type into the Skype chat box, during the conversation, although I do not look at it because I am concentrating on what I want to say.

I corrected your pronunciation while we were speaking but you did not notice, probably for the same reason. It is been my experience over and over that correction of the nature you asked for, has little impact, both in the short and long term.

If you would like your tutor to provide additional comments you just have to ask him or her. That is what I do. As to why you find those two corrections useless, it is just beyond me. You want correction but when some obvious mistake patterns are pointed out, you seem annoyed. Why not look for how "many" is used with verbs, or save more examples of the past tense whenever you see them in your reading.

Igor,

I presume you are writing in what is to you a foreign language. You express yourself well with few errors, but I do not understand what point you are trying to communicate.

Language is about communication.

[Katie]

Once a week, I'd meet with a Spanish tutor in the city, and as we spoke, she'd quietly jot down corrections on a Post It note. If I was interested, I'd take a peak. Otherwise, we'd just continue. I'd save the corrections in SuperMemo.

It helped, but the most satisfying for me was when we'd talk about whatever came to our minds. We became great friends, and our weekly meetings were more like therapy for me. I know my Spanish improved.

My boyfriend, meanwhile, another native speaker, would correct me to no end, and took the fun out of learning for me. He told me that I had a bad attitude, but when you can't finish a single sentence without being corrected, it is the most stifling, depressing feeling in the world. It's not natural. Could you imagine doing this to a child? It'd be abusive. I mean to this degree.

Sometimes it's helpful, kind of like someone telling you you have toilet paper stuck to your shoe. It's something glaring, that you want corrected, so you feel less self-conscious.

But when you correct every single little thing, it kills the spirit.

What's helped me the most is listening and sometimes reading along to conversations with native speakers. When I make a mistake, it's not because I've learned something "wrong." It's that I never learned it at all, and so I'm making up the rules in my own mind. I prefer to memorize how native speakers say things. It takes a lot of listening to do this, but it's much faster than someone pumbling you with corrections all the time.

[Ilya]

Steve, my points are:

a. I encourage the LingQ learners to record their conversations. (For me such recordings are valuable).

b. The value of the recordings would increase if the corrections and comments of the tutors were RECORDED there. (I may not notice or deliberately ignore many of a the corrections during the conversation, but I will make use of them when I play the recordings back).

c. Certain types of the corrections are better communicated by voice.

d. Your general instruction like “Do not disturb the user during the conversation” should be, IMHO, balanced with b and c.

I fully agree with your:

“1)the learner should decide whether he or she wants to be corrected, when and how. The attitude of the learner towards the language, his or her enthusiasm and confidence is more important than any little error correction.
2)correcting is of relatively minor importance in the total process of language acquisition.”

[Igor]

Sorry for my previous "disjointed" comment, I was in a hurry, but honestly I didn't really wanted to discuss about Krashen's claims that "correcting language mistakes is a waste of time", not that I do not agree with that and not that I consider it unimportant but I think you are writing about minor problems instead of focusing on the most important things first and I can't find a post on this blog where you write about those important things and where would be appropriate to comment. For example last time when I wrote to you (that was on the post about graded readers) you said that you "preferred to fight my way through Tolstoy with 40-50% unknown words at LingQ". So tell me now if you are studying languages that way, WHAT ACTUALLY is the difference between your "the linguist" method and old well-known "grammar-translation" method still widely used in most of the classrooms? Nothing, or almost nothing. In grammar-translation you take a short passage and analyze it for new vocabulary and grammar and than do the exercises, in your method you have eliminated the grammar and exercises part and that's it, new method, big deal. No innovations, terrible way of acquiring vocabulary and even worst way of acquiring grammar. If this blog is about promoting a new product that is now free but in future will be sold for big money than fine, just tell me and I'll shut up right now, but if you're really interested in applied linguistics and want to exchange ideas about language learning you should take a crash course in it because now you know almost nothing about second language acquisition theory, you mention Krashen very often, it looks like you like his hypothesis but I assure you that you don't understand it, I have send to you some books in order to familiarize you with it but no results until now. And yes I am writing in what is to me a foreign language, if you want to criticize me about that go ahead I give up, I'm not good enough, it's a language that I acquired, I don't know how and when somewhere in the past and didn't have the time to improve it, I was very busy studying the SLA theory, mostly Krashen and Mason. I apologize if you find me rude, that was not my intention, I have found your website googling the word Krashen, so maybe I expected too much from you and now I am disappointed or maybe this blog was not intended for people like me.

[Chris S]

Igor,
I can't speak for Steve, but I don't think he cares much for SLA. I think he likes culture,humanities and people and sees that the window to it all is through language.

I think very basically his idea is to listening and learn without any pressure and speak when your ready. I think it's simple as that and all he has for proof that it works is that he speaks 9 or 10 languages and enjoys the learning process.

Lingq is more of a way of saying "this is the way I did it" come join me if you want. I don't think he ever made the claim that he was 'linguist' in the scientific sense, buy only in the sense that his 'lengua' has made sense of a lot of languages. Your English is very good by the way, I only knew you were foreign because of your name.

[Igor]

Chris S,
I'm glad that Steve likes culture,humanities and people and sees that the window to it all is through language, also there is nothing wrong if Linq is a way of saying "this is the way I did it" come join me if you want, and let me tell you this story about his 9 or 10 languages: Few days ago I watched my favorite Arsenal playing 4-4 against Liverpool, the man of the match was Andrey Arshavin who scored 4 goals! on Enfild! in the heart of Liverpool! So what do you think, if you ask that same Arshavin to manage Arsenal could he do it just because he's a great player? Not necessarily. In Budapest, June 9, 2003 died Kató Lomb a Hungarian interpreter, translator, language genius and one of the first simultaneous interpreters of the world. Native in Hungarian, she was able to interpret fluently in nine or ten languages (in four of them even without preparation), and she translated technical literature and read belles-lettres in six languages. She was able to understand journalism in further eleven languages. As she put it, altogether she earned money with sixteen languages (Bulgarian, Chinese, Danish, English, French, German, Hebrew, Italian, Japanese, Latin, Polish, Romanian, Russian, Slovak, Spanish, Ukrainian). She learned these languages mostly by self-effort, as an autodidact. Her aims to acquire these languages were most of all practical, to satisfy her interest. My problem is that I thought I saw a successor of Kato, and if Steve cared a little bit more about SLA his word in his 10 languages would have much weight in the world.

[Chris S]

that was a great match... I'm a Liverpool guy though...

[Steve Kaufmann]

Ilya,

I appreciate that different people like to study in different ways. I do not agree with your points a.b.c.d and that is why we do not do it that way at LingQ.

Igor,
I do not agree with your logic.Playing soccer at a professional level requires unique skills and dedication. Language learning is available to anyone. A good scorer in pro soccer is a model for aspiring soccer players, but not necessarily a model for people who want to manage. I do not say "do it may way because I am a star". I say do it my way because it has worked for me and many others. None of this qualifies me as a manager of a school.

Your interpretation of my method is erroneous but I am not going to rehash all of the any things I have said here about input based learning. I have little interest in linguistics or the formalized study of SLA. I have read lots of books on the subject, and have heard teachers debate various theories, and it all strikes me as sterile. The reality of language learning is much more straight forward, in my view, as I have explained in great detail here over several years.

If you find this space interesting fine, but I would suggest you be less dogmatic and shorten your posts a little.

[victor]

Igor, after googling for "Krashen" google for "politeness" too.

[Igor]

Yeah right, I understand now, you have answered my crucial question, good luck with your blog.

[Igor]

P.S.

and you should google for "Dialoguing", you're proud representative of it.

[staphl]

Hi there!

My native language is German and I consider myself fluent in English, I get along in French, Spanish, Turkish and Georgian, I learned how to translate Caesar and I can produce a few phrases in Russian.

In my humble opinion, uninvited correction of the spoken word is just a nuisance. When I have a question about something, I will ask.
But when I am trying to speak about something else, I surely don't want anybody (let's say, my Georgian husband ;-)) interrupt me and try to make me pronounce a word correctly.

I actually believe that, while passive language acquisition is about listening, reading, studying, the active use of a language begins most of all with the courage to make a lot of mistakes.
Maybe for other people it works differently, but for me the most rewarding part of language learning is communication with other people, no matter how basic.
When I see that I can find my way, buy everything I need and speak with taxi drivers about the price, I already feel that I am a little bit independent and that makes me so proud and happy that I want to go on in learning the language.

I also don't usually correct my language tandem partner without being asked. She doesn't make a lot of mistakes and if it's just the wrong word order, I trust her to understand it by herself eventually.

With my son though, I don't say something like "No, that's wrong", but if he says something the wrong way, I repeat it the right way.
He is very small, not even one year and a half and we are just learning the words for animals, food, body parts, some numbers and so on.
For example, he will ask: "Kokoli?", I will answer "Schokolade?" (chocolate) and then maybe he will get some. I don't think that that really counts as correcting.

[sjoerd leferink op reinink]

I disagree on 1, 2, 3 and 4. First of all, if you speak my native language then I'm going to correct you. After all you are trying to speak my language. I cannot hear you making major faults when I can- or may not correct them. I do mean major fauls, no minor ones. I don't care if it's good for the leaner or not. Maybe that makes me rude...but it's my mothertongue so I'm going to interrupt and correct even when the leaner doesn't experience it as friendly. #2 If I'm am going to correct the learning, then it's my choice whether if I keep it to mininum or not, he/ she just has to take that. After all it's my mothertongue. #3 the comparison with written or spoken is not relevant, because you always correct someone different on writing skills than on speaking skills. Little grammatical rules doesn't play a role in speaking sessions. Writing is different than speaking. I think that is the same in every language. #4 Why after the conversation? He/ She should speak it good. If they don't, I'll tell them, in the middle of the conversation or afterwards. I thinks you lay too much emphasis on elevating the learner. If I understand you, then I think you lay too much importance on the learner and too little on the teacher. Like the learner is much better than the teacher....if I'm wrong correct me...

gr. sjoerd

ps. I agree with #5

[Steve Kaufmann]

Let's just say sjoerd, that I completely disagree and would not have you as a teacher, nor would I speak to you in your native language if I were learning it.

Assuming we are all adults, the learner is in charge. It is his or her choice to learn or not to learn. The teacher is merely a resource, to be used as the learner sees fit.

You remind me of the famous saying of one great leader "The beatings will continue until morale improves."

[sjoerd]

The teacher is not merely an instrument/ resource for the learner's own interest. Why is the learner in charge? He/ She is learning the language of other people/ countries? How can you say the teacher is nothing but an instrument/ resource? It's maybe an offense to tutors...

"The beatings will continue until morale improves."

What a quote...do you insinuate I'm some sort of a gothic barbarian who likes to punish people? Come on...

gr. sjoerd

[sjoerd]

I think I understand what you mean, but I think it has to come from both sides and not just from one side. If you don't want me as teacher or even want to speak to me in my language and create a ""cordon sanitaire around me...I don't care. I don't really have a lot of time anyway...

gr. sjoerd

[Steve Kaufmann]

No you do not understand. The learner is in charge. If the learner asks to be corrected that is one thing. No one, teacher or native speaker, has the right to impose corrections on a learner. The learner can just walk away. The teacher has also the right to walk away, and decline a student, but the teacher does not have the right to impose him or herself on the learner. It is all about the learner.

[Friedemann]

sjoerd,

I agree 100% with Steve in that motivation and encouragement are key to language learning. The learner, at least the adult learner should be allowed to decide if he or she wants her mistakes corrected. To force corrections upon someone is counterproductive and rude, I think. What would you say if I were to lecture you about all the mistakes you made in your initial post above?

I like being corrected in my language studies, but again, that is my choice.

Friedemann

[sjoerd]

I think I'll have to revise my view. I think I drank too much last night when I wrote that above. I've made this mistake more often...

I agree on 1, but what should you do, as a tutor, when your student constantly is making major faults and it doesn't make sense at all? I agree on #2 and #3. #4 why just a few comments? Maybe you should give a few comments to beginner learners, but why can't you give more comments on the more advanced students? I agree on #5.

gr. sjoerd

[Ilya]

sjoerd: "I drank too much last night when I wrote that above. I've made this mistake more often..."

Steve: "In fact mistakes are good, we need to be made aware of them, but in the right way."

Drinking over Skype. Free 30 day trial.

[Karen]

Hi Steve,

I have a question for all of you. I teach English in Canada to immigrants. Some of the students in my classes have been in Canada for more than 5 years. They were taught by a teacher who used a lot of talking and writing but very little grammar instruction and, I think, not much correction. After all this time, their English is still very poor. One student speaks very broken English and understands just basic conversation. The others (8-12 years in Canada) understand fairly well but their speech is very difficult to understand because of their incorrect pronunciation and poor sentence structure. When they speak quickly I can't understand them at all.

Any ideas? How can I help them break the bad language habits they've formed? It isn't even really their fault, some of them are very diligent students.

[Steve Kaufmann]

Karen,

It would be helpful to know how many students are in your class and how many hours they are in class, how literate they are, what they are interested in etc.

I do not think you will like my advice.

First of all their language skills, after 5 to 12 years in Canada, are entirely their responsibility in my view. Until they recognize that they will not improve. That is where you need to begin. Their attitude towards the language, their determination to improve, their belief that they can learn to speak fluently, these are the keys to their success. Your first goal is to try to inspire this kind of attitude in them.

Second, what they do outside of the classroom is much more important than what they do in the classroom. There is considerable research to indicate that correction and grammar explanation do not do much, so I doubt that the lack of it, over a period of 5 years, is the reason their
language skills are poor.

In my view, they need to listen a lot, several hours a day, to English. The best way is to load content of interest to them on an MP3 player and always carry it with them. Listen every spare moment, while doing chores, while commuting etc..They should choose content they like, where they like the voice, and listen. They need to listen to the same content over and over imitating. They need to notice words and phrases, save them, review them and imitate them. When they do write and speak their corrected phrases should be added to this list.

They should also take every opportunity to watch TV listen to radio, read the newspaper and talk to people around them in English, not their native language, at least for a concentrated period of 6 months or more.

I would suggest they join LingQ. You should join LingQ too, and work on some other language. That way you would understand their situation.