Correction... how useful...?

[Le Baron]

Student:
1) When the teacher says something in a way that you wouldn’t, imitate her.

As a student, this is something I've tried to do. Or when speaking with a native asking the question: 'how would you say it?' The question because there is sometimes a tendency for a native to simplify in an attempt to 'help' you.

[Irena]

I generally prefer lots of corrections from a teacher. As a beginner, I prioritize accuracy over fluency, and once I'm advanced, I won't make that many mistakes anyway. So, if I get something wrong, I'm generally okay with being stopped and corrected. (Though one of my Italki teachers would type corrections into Skype chat, without actually interrupting me. But that functions as an interruption of sorts anyway, assuming I'm paying attention to the chat, which I normally am.) And if I want background information, I'll ask for it.

The reason I prioritize accuracy over fluency is because if I start learning a language, my goal is to get to a high level (I may fail to get there, but then I'll consider that a failure), and for that, I think it's more important to reduce the number of fossilized errors than it is to increase fluency as quickly as possible. But that's just a personal preference, and it's probably true that even I could get too much of a good thing (corrections from a teacher).

Now, this is for teachers! I wouldn't want to be corrected in random interactions with native speakers, except on rare occasion.

[Le Baron]

On the question of being corrected by either teachers or in conversation, I overheard a conversation yesterday when I was in the city. It was very agreeable weather and I was at an outside table at a 'cafe theatre'. There was a young fellow whose nationality I couldn't determine, I'd guess at Spanish or Italian, and he was speaking English to someone else. An old Dutch geezer dressed like a youth! He was telling him about how he'd been in Utrecht for a while and was planning to go somewhere else for three or six months.

This is a particular sort of case because it is two non-native speakers of English, though it is worth looking at for the phenomenon of why certain incorrect formulations persist in foreign English and perhaps any other language. Let's say he was Italian, he kept using two structures you hear a lot here: 'I was' instead of 'I have been' and 'I am going/eating/doing' etc instead of 'I go/eat/do'. So he told the old geezer: 'I was here since January 2022...' Which of course made the cogs in my head start to gnash and smoke. The Dutch fellow didn't bat an eyelid because he probably says this himself, or knows 9/10 people say it and it's 'normal'. Then the 'Italian' said: 'I am going to the school every day since then...', which I assume is the university or some other college. Again this is just accepted and also doesn't actually impede comprehension probably at all, even if someone knows it's not correct.

Just as an aside, it always baffled me why those formulations are used here in NL. In Dutch if you say e.g. 'I go to the gym twice a week..' it would be: 'ik ga twee keer per week naar de sportschool...' Ik ga, which is exactly the same as I go. Yet when translating it to English the Dutch (and lots of foreign) speakers commonly choose 'I am going...' instead. It seems because Dutch and some other languages use only the one form for 'I go/I am going' there's a tendency to assume that the correct translation into English is always the second one.

Obviously I can't know the background of every learner. Whether they learned English 'autodidactically' or picked it up haphazardly on top of the usual basis of school English, or whether they followed a structured course or whatever. It seems the 'real life' free conversation arena, where everyone really achieves functional use, has a stronger impact upon all this. It makes sense and accounts for every type of phenomenon such as accent, word and phrase memes in vernacular speech including baked-in 'errors' (I could care less etc), common word choices. So I wonder if such people were ever corrected, or if the corrections they have received were ever strong enough to override this problem. A native speaker of any language doesn't require these types of correction, or as much, because in general the correct examples of basic speech are intact in the first language you ever learn. Even if you don't get them first time it's quickly reinforced around you by peers and schooling and life in general, in a language in which you are already functional. Mainly on the basis that this is your native language (godammit!) and you ought to be getting it right to this extent!

Learning an L2 is different. The time scale is far shorter for what is expected. It isn't a case of 24/7 exposure over a generational lifetime, examples from the bottom upwards, but a transference between two (or more) forms of already working language which may also even be poles apart. So considering that the average learner here is always looking for 'learning efficiency', why wouldn't they accept and use correction? It's clear to me that simple exposure and listening, even over a long period, and self-correction which is limited, actually doesn't iron-out sometimes rather egregious errors. I would reformulate the question to: corrections...can you ever really avoid them?

There's also the obvious fact that the demand for 'perfection' or near perfection, while that is a legitimate personal goal, probably shouldn't be such a strict requirement for L2 acquisition. I could have had a perfectly good and functional discussion with that 'Italian' fellow and have done so with many like him. What kind of balance do we need between correction and good functioning? Some people probably just need correction more than others? Such as if you want to write (academically maybe) in that language or exude a sense of authority.

[Irena]

@Le Baron

If you corrected the Italian fellow, he would probably just conclude that you're an @$$hole. But seriously, what's the point?! If you can understand him, then you can understand him. Now, if you were his teacher, that would be a different matter. And it's different in a classroom setting (where the teacher must set the tone), and in a one-on-one setting. For the latter, different learners want different things. Some just want to be understood. (You know how Steve Kaufmann says that it's no problem to mix in various Slavic languages when speaking to Polish speakers, as long as it's Polish enough for them to understand you? Well, the Italian fellow is miles ahead of Steve Kaufmann! ) Others are after a high level of correctness. It helps if the teacher and learner are on the same page.

[Irena]

And then there's the related question of how exactly you get rid of fossilized errors once you've already achieved a high level of fluency. I actually have no idea, even though this is precisely what I'd like to do with my French. My French is super fluent (or at least it was super fluent until I put it on the back burner a few years ago in order to focus on Czech). I scored 90% on the speaking part of DALF-C1 (my other subscores were much lower), and I guess that was due to my high level of fluency. Oh, but those errors... Far too many errors! I guess some things were either never explained to me, or they were explained but went over my head (I started learning French as a child). I keep telling myself I should do something about this, but I'm not entirely sure what.

[Cainntear]

But seriously, what's the point?! If you can understand him, then you can understand him. Now, if you were his teacher, that would be a different matter. And it's different in a classroom setting (where the teacher must set the tone), and in a one-on-one setting. For the latter, different learners want different things. Some just want to be understood.

As I think I already said, a teacher is in poor position to deal with "comprehensibility" as a criterion or learning goal. Teachers are exposed to more learner errors than the average speaker, so a teacher's ability to understand is significantly higher than the average speaker.

Teachers will often avoid correction on the grounds that "I understand it so it's not a problem" but are generally ignorant of the fact that other people might not.

Teachers are in no position to say that an error won't impeded comprehension when said to non-teachers, so allowing a learner to repeat their personal common errors is poor practice.

[Cainntear]

I generally prefer lots of corrections from a teacher. As a beginner, I prioritize accuracy over fluency, and once I'm advanced, I won't make that many mistakes anyway.

Well exactly. Allowing students to make errors early on by saying "it's not important, they'll still be understood" is the wrong approach because it ignores the fact that it doesn't take as much effort to learn something right in the first place as to remove the error late.

I don't think this is a matter for personal preference at all, because learning the correct form from day one makes the learning on day 2 easier, which makes the learning on day 3 easier and so on, because every language has its own core elements from which everything else grows.

...but...

It's not really a choice between fluency and accuracy. Michel Thomas's course result in high fluency and high accuracy. Yes, they're only introductory, and no, he didn't set out any advice on how to continue once you'd finished his courses.

They are a long way from perfect in many ways, but as far as I can see, they are very good proof of how much better teaching can be.

[Irena]

I generally prefer lots of corrections from a teacher. As a beginner, I prioritize accuracy over fluency, and once I'm advanced, I won't make that many mistakes anyway.

Well exactly. Allowing students to make errors early on by saying "it's not important, they'll still be understood" is the wrong approach because it ignores the fact that it doesn't take as much effort to learn something right in the first place as to remove the error late.

I don't think this is a matter for personal preference at all, because learning the correct form from day one makes the learning on day 2 easier, which makes the learning on day 3 easier and so on, because every language has its own core elements from which everything else grows.

Well, I hate to argue against my very own preferences, but: it is a preference. If your goal is to get to an advanced level in your target language, then yes, I do think you will just sabotage yourself by not getting your errors corrected early on, for precisely the reason that you said: it's easier to learn something properly early on than to fix it later. But. Many people aren't in the business of getting to an advanced level. They just want to communicate, often at only a relatively basic level. Le Baron described one perfectly successful interaction that took place in imperfect English. What he described is in no way extreme. Think about people who have multi-language conversations (one person speaks in one language, the other one answers in another). Assuming the languages are close enough (e.g. Czech and Slovak), it works. And even if they aren't that close (e.g. Czech and Polish), you can still kinda make it work, especially if you throw in some words from the other language (while most of the words words, plus the entire syntax, come from your own language). Is it going to work great? Well, no. However, a non-trivial amount of communication will take place. Now imagine how much better it gets if you actually sorta learn the language, even with a bunch of mistakes. Polish speakers will understand a native Czech speaker who spoke bad Polish better than they will understand a native Czech speaker who just spoke Czech with some Polish words thrown in.

It's not really a choice between fluency and accuracy. Michel Thomas's course result in high fluency and high accuracy. Yes, they're only introductory, and no, he didn't set out any advice on how to continue once you'd finished his courses.

Eh. I don't know. Do you just blurt out the first thing that comes to mind, or do you monitor what you say? If the former, then you're privileging fluency over accuracy, and if the latter, then it's the other way around. Eventually (as you become advanced), you can have both. Early on, though? I rather doubt it.

[Cainntear]

Well, I hate to argue against my very own preferences, but: it is a preference. If your goal is to get to an advanced level in your target language, then yes, I do think you will just sabotage yourself by not getting your errors corrected early on, for precisely the reason that you said: it's easier to learn something properly early on than to fix it later. But. Many people aren't in the business of getting to an advanced level. They just want to communicate, often at only a relatively basic level.

As I say, getting accuracy right from day one makes learning a language quicker.
Teaching well is better for students who value fluency over accuracy and for students who value accuracy over fluency.

Teaching well is neutral to student preferences. A student who values fluency over accuracy will benefit because being able to effortlessly get the syntax write means they can speak faster and be more comprehensible.

It's not really a choice between fluency and accuracy. Michel Thomas's course result in high fluency and high accuracy. Yes, they're only introductory, and no, he didn't set out any advice on how to continue once you'd finished his courses.

Eh. I don't know.

Of course you don't know, because you haven't taken a Michel Thomas course.

Do you just blurt out the first thing that comes to mind, or do you monitor what you say? If the former, then you're privileging fluency over accuracy, and if the latter, then it's the other way around. Eventually (as you become advanced), you can have both. Early on, though? I rather doubt it.

There was a time when I was halfway through that my sister started faking kicks to my face on the way home from a night out with her Spanish friends. I made a slight error when I spontaneoulsy said "don't do that" in Spanish, but I think it was just because by that point I was more comfortable with Usted than tú, particularly in the imperative mood. My sister was insulted by the slip but the Spanish guy with us was astounded, because complete beginners are normally very restricted in the conjugations they can use for tense and aspect. I had higher fluency and higher accuracy than any other course would have given me.
Thomas wasn't forcing memorisation of rules, which meant that being accurate didn't mean consciously analysing the rules to ensure correctness, and because there was no analysing to be done, there was no need to stop analysis in order to speak at speed. i.e. There was no need to sacrifice accuracy in order to increase fluency.

Teaching that makes you choose between accuracy and fluency is very, very far from suboptimal, because it accuracy practice encourages you to treat accuracy as declarative knowledge while fluency practice doesn't encourage the transferral from declarative to procedural, because it reinforces the use of procedural habits from L1 transference in the new L2.

The fact that there are now flaws in my Spanish isn't MT's fault.

[Iversen]

Teaching well is neutral to student preferences. A student who values fluency over accuracy will benefit because being able to effortlessly get the syntax write means they can speak faster and be more comprehensible.

Student preferences undermine the idea that there is one single thing which could be called "teaching well". As a teacher the best you can hope for is that your teaching style is relatively OK for the majority of your pupils/students while still leaving room for those who for some reason can't accept it. And that's where I was lucky: I had teachers that did their thing at the blackboard, and then I just studied for myself at my desk unless they disturbed me.

If your goal is to get to an advanced level in your target language, then yes, I do think you will just sabotage yourself by not getting your errors corrected early on, for precisely the reason that you said: it's easier to learn something properly early on than to fix it later. But. Many people aren't in the business of getting to an advanced level. They just want to communicate, often at only a relatively basic level.

When I start a language (which I have done several times) my goal would always be to reach a level where I can express myself fluently without making too many errors. OK, in some cases I settle for just learning a language or dialect passively because I can't see any chance of ever needing to use it actively (or because of a scarity of study materials), but that's an exception - normally I reach at least a level where I can write in my languages - and then I can of course also read them. And I can speak a decent share of languages well enough to have hourlong conversations in them or do a lecture - how much more do I need?

Along the way I have found out the fossile errors is a myth: the thing that has fossilized is the complacency of the learner. You are in a better position to correct an error when you already know a language reasonably well, but precisely at that point you may be too satisfied with what you have got to bother.

I have mentioned my separation between intensive studies and extensive studies several times here. The intensive ones is where I collect facts, and that's a slow process where I have to be meticuluous about things. The extensive activities serve to fixate all that knowledge and to make passive knowledge active, and there it would be counterproductive to be too perfectionistic. When I try to think in my mediocre languages I run into lacunes all the time, but if I felt I had to look all those words and expressions and constructions up I would never get anywhere. But with time the holes get fewer and my fluency improves, and with a bit of perseverance I may get to a level where I also get complacent and accept the status quo.