Correction... how useful...?

[Cainntear]

Somebody reopened an old conversation on Vladimir's videos (I find the channel name so thoroughly insulting I won't repeat it!!!). In the middle, the conversation slipped onto the topic of correction, and I thought that was worth bringing out into a new topic:

…

When it comes to language learning, I do everything by myself except converse. And so the only job of my teachers is to converse with me. I don’t like corrections from tutors because it detracts from the purpose of the lesson – getting comfortable in the language. Because I learn everything on my own, I hear many of my own errors when I speak. The more I speak, the fewer my errors. Correcting me just distracts me and makes me less fluid. This is especially true in the beginning. I feel I’ve got so many balls in the air, I don’t need someone adding more. The better I get, the more open I am to correction.
…
Here is a generic note that I send to all my teachers the same time I send a lesson request:

Hi {Teacher X}. I'm an intermediate student. I'd like a 100% {Language X} conversation with no corrections and no fixed topics if possible. Thanks

99% of them agree to this, and the result is a low stress conversation that lets me become more comfortable. We talk about whatever comes to mind. I’ve done thousands of these, and improvement is really fast when I do the following: During the conversation, write down all vocabulary/grammar points that the teacher uses but I don’t understand AND all the vocabulary/grammar points that I want to use but don’t know. After the lesson, I memorize these items, then put them in Anki. Before the next class, I do the Anki reviews. If possible, do a 60 min class every day or at least a few times a week, and repeat this exercise.

I found this post quite interesting because it is very much the opposite of what I do with tutors. I’m not saying that there is any thing wrong with the suggestions in the post. It’s the results that count. It only goes to say that there are many ways to skin a cat.

In my approach correction is paramount because I believe this is the key to accuracy. By accuracy I mean two things. On the one hand there is pronunciation and grammatical correctness : I want to say things the best way possible and I want to make sure I use the right word forms in the proper order.

On the other hand I want to use the most appropriate vocabulary for the context. This is the discourse level. For this I carefully listen to how native speakers formulate their phrases and I try to imitate them.

I therefore crave correction because correction means that something is wrong or could be improved. The problem here is that for some people, rightly or wrongly, being corrected means stopping the flow of the conversation and preventing fluidity from developing. My fear is that if I’m not corrected I will develop and solidify bad habits or fossilizations.

For me the question is how to be corrected. There are a couple of different situations here. Number one, let’s say I use the wrong word form. I say neuron in Spanish. The tutor stops me and says neurona. Number two, I am not sure of the form to use. Let’s say that I think I should use the imperfect subjunctive form of a verb in Spanish but I’m not sure of the form to use. So I stop dead in my tracks and I explore various forms until I find the right one.

In both cases what I like to do is repeat the whole corrected sentence or passage a couple of times until I get the whole thing right fluently. Of course I take notes to review later on, and I will make a point of deliberately using the corrected form and variations as soon as possible to reinforce the learning process.

I don’t see this as preventing the acquisition of fluidity. Quite the contrary, I see this as enhancing fluidity in that it is conducive to acquiring accurate fluidity whereby you develop true confidence in speaking well because you are totally sure that what comes out of your mouth is correct.

I know that this sounds more appropriate at advanced levels of proficiency but I also believe that even in early stages one can nip bad habits in the bud without necessarily making speaking laborious and boring.

I thought this was quite interesting, because I'm not sure whether they're agreeing or disagreeing with each other, and I kind of agree with both, but more with s_allard than leosmith.

There's a bit of a thing about correction in mainstream teaching that is pretty problematic in my view.

Delayed correction (note errors down and give them attention at the end) has become essentially "received wisdom" at this point. My own experience as a learner has been that late correction basically repeats a rule that I already consciously knew, but was failing to apply in fluid speech. I know I've seen looks of frustration on learner's faces during correction sessions and I'm pretty certain I've even had students respond verbally with "I know" or "I know that" showing clear frustration.

Delayed correction can't work, because the at the time you are giving the correction, the learner isn't accessing or activating the system for spontaneous output -- all you are doing by giving delayed correction is teaching declarative knowledge of language and not procedural knowledge, and this is particularly self-deceptive when the teacher/course proclaims "no grammar" because the philosophy behind no grammar is that of there being no transference from declarative memory to procedural memory. Delayed correction relies on strong transference from declarative to procedural memory.

But what about correction during spontaneous production? If you correct every error, that breaks the flow of the conversation, right? Right.

Does that mean "no correction" is a good philosophy? No.

My favourite courses are by Michel Thomas. MT corrects all errors. Why is that not breaking the flow? Because (a) this isn't a conversation so there isn't a flow; and (b) he controls the class to such an extent that very limited numbers of errors are possible, so his correction is normally a very trivial distraction.

I have often been in a classroom as a student and I've understood my classmate and known what the correction should be, but the teacher has misunderstood and given the wrong (mis)correction.

This is a contrived example, because I can't remember the real circumstances, but if I say "I did it tomorrow", do I mean "I did it yesterday" or "I will do it tomorrow"? We do not know, so a teacher that jumps in to correct it without asking has a 50-50 chance of saying the right thing. That's clearly not good enough. But that doesn't mean that all correction needs to be avoided.

I had a friend from the local Spanish exchange who asked me for correction-based English lessons (and IIRC he was going out with an English teacher at the time, and just felt that asking her for lessons would be a quick way to sabotage their relationship).

He wanted me to stop him and correct him. Did I correct every mistake? Of course not. I listened intently and determined which errors were most important, and which errors could be cured by addressing completely different errors -- i.e. I only corrected what I identified as "root errors" and then attacked "consequential errors" by drawing his attention back to the (now corrected) root error. I also moved very subtly from drawing his conscious attention to an error by explaining it in a verbose manner, to giving just enough of a prompt to trigger semi-spontaneous recall of the rule. Eventually (well I say "eventually" -- he wasn't my student for more than a couple of months, one hour a week) I got to the point where I could indicate to him that he had made an error just by a slight change in my physical attitude -- maybe wrinkling my brow or sitting up straighter very suddenly -- and that was enough for him to self-correct in a way that seemed almost spontaneous to him.

I have a vague recollection of being in a pub or at a party and me commenting (to someone else who he had commented on me being a good teacher to) that I could correct this guy just by raising an eyebrow. He wasn't convinced, because he remembered the more conscious teaching. So the next time he made one of his more frequent errors, I raised an eyebrow, and his spontaneous laughter told me he'd completely got the point.

He's probably the only guy I've ever had the opportunity to do that with, and thinking back on it, I'm immensely proud (he was telling me that his coworkers in the restaurant he'd been working in for a year noticed how much his English had improved in a month).

[Le Baron]

Correction works. Especially for repetitive errors, corrected at the time and then re-employed during exchange as examples to learn from so that it is made into a habit. Also saying why it is incorrect.

However I don't believe in one-size-fits-all (I am after all a bespoke tailor). The student's temperament matters and so does the teacher's tone. I don't like Michel Thomas's tone in his corrections, because he sounds impatient, even if he is or isn't, and this can have a bad blockage effect on the learner who feels more foolish than assisted.

Obviously it matters what is corrected. Mere 'style' where there are no real errors shouldn't be corrected, that's just pettifoggery. Wrong use of prepositions should be and in my experience people made of aware of it tend to be very grateful and do fix their errors. I've done this in English, French and Dutch and seen error adjustment occur.

You can't know what you don't know and often don't see errors until they are pointed out, so the view that they will iron themselves out or you can always 'teach yourself', autodidact or not, seems to me unfounded. I've met autodidacts with multiple languages learning French and Dutch who make glaring mistakes. They are habits where no error is even assumed because of not knowing.

[Iversen]

It's complicated ... If I'm in the middle of a conversation then I definitely don't want corrections because they break my concentration, and since I don't take classes anymore there are few options for correcting me during intensive study sessions. In an old thread a related question was raised, and there I stated that I preferred corrections which could be applied to a large number of words or constructions. Example: I accidentally found out that Finnish ä is pronounced more like a flat a than a German or Swedish ä with Umlaut, and then I could change my overall phonological model for that language in one fell swoop, hurrray.

Corrections to one word or construction are in principle also welcome, but I think they would have most effect if delivered in an unobtrusive way right after the error was committed (like for instance by quietly saying "neurona" if your hear an erroneous "neuron" in Spanish) and preferably during intensive study sessions, but since I don't employ personal trainers or teachers that possibility is out. The corrections that come days or weeks later are not wasted, but they range at the same level as my own random discoveries during extensive activities - like for instance that it is нейрон ('nejron') in Russian.

[Le Baron]

If I'm in the middle of a conversation then I definitely don't want corrections because they break my concentration,

The way to do it (the way I do it) is to make it part of my reply, where I (re)use the part of the sentence with the error, but correct it. This very often works as the learners then correct themselves and there is no overt 'correction' to break flow or make anyone feel inadequate.

[Iversen]

Oh, you are just too fast! I was inserting a reference to the ''neuron correction from the first post into my own answer while you posted your last answer, but hen I saw that the correction was delivered by a tutor - so now I have to change it again. Making the correction by saying the right thing (maybe with a raised eyebrow) is of course the optimal way of being non-obtrusive.

[Beli Tsar]

If I'm in the middle of a conversation then I definitely don't want corrections because they break my concentration,

The way to do it (the way I do it) is to make it part of my reply, where I (re)use the part of the sentence with the error, but correct it. This very often works as the learners then correct themselves and there is no overt 'correction' to break flow or make anyone feel inadequate.

Which - probably not coincidentally - is the main way parents correct their children's speech.
It can be subtle, not crush the ego, and provide a good model for better speech all in one.

[leosmith]

Delayed correction can't work

Rather than "can't work", don't you mean something like "may not work as well as you expect it to"? Otherwise, it sounds like you are saying people who do not receive correction, or receive only indirect correction, do not improve. Or maybe you think they only improve due to other reasons.

[Iversen]

While we wait for Cainntear's answer I'd like to comment on leosmith's question by adding one aspect to my earlier answer, namely: what would it take to make delayed corrections work? And here I would compare them to random things I see on TV or on the internet or in written sources.

If I'm told that a certain word doesn't mean what I thought it meant or that a pronoun should be before an adverb and not after or that a certain idiom isn't used correctly then it will range at the same level as random information I run into on the internet or on my telly UNLESS I still remember the original context clearly and feel a responsability for it. OK, there may be an added incentive to take it seriously because somebody took pains to correct you, but I would just accept the correction at face value and move on.

But sometimes I'm in the mood to investigate the correction, and then I might look the word or construction up to find more information, and in a few cases I might then find that the correction in itself was dubious. For instance I was once told by another Greek learner during a travel that I had used a wrong preposition in the Greek translation of "a ticket to Athens". OK, when I came home I looked the expression up on Google and found that MY choice was by far the most common one. We have also seen discussions where a native speaker says that so and so, and another native speaker says that it ain't like that where she/he lives. But even if the correction seems to be valid the extra effort of checking it makes it almost certain not only that I remember it, but also that I learn a lot more in the process.

So to teach me something a correction should either come while I still struggle in the aftermath of having produced something OR it should be of a kind that makes me wonder whether the correcion was warranted - or lures me into studying the topic further for some other reason. It also matters whether you get a whole lot of corrections or just a few. I would guess that getting too many corrections in one go erodes your motivation (and possibly selfconfidence). So if a teacher practically rewrites all you had done then it doesn't teach you anything because you just give up. Even if almost everything really IS rotten then it would be more constructive just to hint at the general low level and then point to maybe four or five things which could be remedied here and now.

By the way, during my French studies in the 70s I once opted out of the conversation classes by a certain prof while she simply refused to accept that I wanted to say that a baby couldn't be used for anything else than howling. She ascribed it to my (in her opinion) low level of French and tried to reformulate it into something meaning other more palatable. I simply reject all corrections of the type "what you really want to say is...".

[Cainntear]

Correction works. Especially for repetitive errors, corrected at the time and then re-employed during exchange as examples to learn from so that it is made into a habit. Also saying why it is incorrect.
...
Obviously it matters what is corrected.

To take that one step further, it matters why it is being corrected. It probably seems too obvious too need stating, but my experience says it isn't: many teachers will give a full explanation of the error multiple times. A good teacher, an experienced teacher is likely to intuitively reduce the amount of explanation over time, but they're not consciously aware they're doing it, which means the explicit rule they think they're teaching by, and that they pass on to inexperience teachers, is "explain every error".

Mere 'style' where there are no real errors shouldn't be corrected, that's just pettifoggery.

Actually, I suspect that's a consequence of "correct everything" philosophy -- most speakers aren't good at correcting actual errors and their attention is caught by anything that "sounds weird" or which their own primary school teachers taught them not to use.

The problem is really that there isn't enough talk about correcting errors other than the "do correct errors" vs "don't correct errors" opposition. What I'm trying to do with this thread is get people thinking and talking about what correcting errors actually means, so thank you for doing so.

Wrong use of prepositions should be and in my experience people made of aware of it tend to be very grateful and do fix their errors. I've done this in English, French and Dutch and seen error adjustment occur.

Yes, prepositions are often neglected, and it is often ignored that a preposition error is very much prime for correction. I think the reason these are under-corrected is that there isn't really a good explanation, though. Language A says "through" with a certain verb, language B uses a preposition that normally means "over" -- no real why can be found, all there is (usually*) is the fact that it's idiomatic. (* if people ask me why we go "on" a plane in English when we're stuck in a pressurised cylinder, I tend to draw their attention to "on the bus" and point out that early buses were fancy carriages with no roof, so "in" wouldn't have made sense, and "on" seems to have gotten fixed as the public transport way of speaking because of that.)

[rdearman]

Having been a victim of correction myself I am definitely on the side of leosmith. At least as far as language exchange goes. Most natives do not know how or what to correct so they stop you every other word and it is annoying. I don't often drop language partners but that has been the cause of most of the dropping I have done. While it might be different for correction by a teacher interruptions would still annoy me. So it had better be important

Personally I only want the correction of systematic errors. Errors which I have made 3 or 4 times in a row then yes correct me but don't stop me for every single error.