On the very questionable value of error correction at intermediate level (as commonly practised)

We are language teachers for a reason: language learning is important to us; it is one of our primary foci in life, hence we pay attention to corrections, they are salient to us.

However, this may not be the case for most of our students. Many of the corrections we feed them, unless we make them very distinctive and they are easy to remember, will be forgotten at a ridiculously high rate (nearly 70% of the information decaying from Long-Term Memory within 9 hours from processing it).

What baffles me is how often some teachers seem to forget  that grammar correction is a form of instruction. Like instruction it does one or more of the following:

  1. it teaches
  2. it re-teaches (clarifies, expands, etc.)
  3. it consolidates

Would you ever be able to do any of the 3 things above effectively through a simple correction or sets of corrections in the margin of an essay? I wish grammar instruction was that simple!

Also, would you expect your average students to AUTOMATISE the faulty items thanks to your correction in the margin, by self-correcting or by having one, two or even three remedial error correction sessions devoted to it after marking an essay? Obviously not.

Would you teach several grammar structures in one lesson, as way too many teachers do during post-essay remedial sessions, by providing a PPT with one slide for each error they found in their students essay? No, of course, because they will cause divided attention in most – not all – of the students.

Would you expect most of your students to transfer the corrective information in your feedback to the next essay they write? Of course not, unless you heavily prompt them before they write the next essay through a reminder or checklist (like, I must admit, I often do).

Would you expect most of your students to learn grammar independently? To do the automatisation work that grammar learning requires on their own through hours and hours of self-initiated practice? Maybe from the highly motivated few.

Finally, would you expect your child, when you tell them they have done something wrong to take it onboard straight away, even if they know you are right and never do it again? And if they do that naughty thing all the time, will they stop doing it unless you promise some sort of punishment they fear or some enticing reward?

Mistakes are part and parcel of the process of learning a language. Many of them are developmental and will disappear at some point through much practice and judiciously administered negative feedback. Many others are due to the fact that you are asking your students to write too much too soon thereby causing cognitive overload and processing inefficiency. Others still because we may have not taught and scaffolded a given structure effectively.

A friend of mine once compared error correction to teaching vocabulary through flashcards; you would not teach a word or phrase by showing the same flashcard once or twice using the same technique, she said. Rather, you would show that flashcard many times over; play a range of games with it; alternate choral and individual repetition, etc.

For correction to truly work with MOST of your students you have to re-teach a problematic item over and over again extensively through masses of exposure to aural and written input and lots of productive practice across a wide range of contexts. The results will take a long time before they actually show because language learning is about the invisible, inaudible, intangible process that unfolds inside the students’ heads day in day out and not the correct sentence in the next essay which gives us the (often false) reassurance that they have ‘got’ it.

Language acquisition is not the linear process that many teachers take it to be. Errors may disappear, then reappear and finally disappear again following a U-shaped developmental pattern than has been documented my much research and is regulated by complex cognitive processes and constraints that are not as yet fully understood.  As teachers, it is crucial we recognize this and shy away from short-term, intensive remedial interventions that, as research clearly shows, do not work and imply a very simplistic view of language acquisition.

In conclusion, in order to improve writing output one needs to do a lot of modelling through comprehensible input and guided practice. Corrective feedback, better if personalised and provided in one-to one-sessions, can indeed assist in the process, but only if it is sustained over the long term and may work only if the student is motivated and perseverant and willing to work independently.

But even so, it will be your modelling and scaffolded practice which will ultimately do the trick. Modelling through Listening and Reading, of course, doesn’t mean giving a text and a few comprehension questions, but getting the students to process what they read and listen to in detail (e.g. through narrow reading tasks or TPRS-like circling) so that they may notice the key grammar structures and/or lexical patterns you aim to teach or consolidate.

Corrections worked for us language teachers and may work for the keenest of our students because they value our corrections and love languages. But what about the average student?

You know my conclusions already if you are a regular reader of mine: better investing your time in planning and resourcing your teaching more effectively.



Are they truly ready to write that essay? – Challenges and solutions for your struggling intermediate student-writers (Part 1)



So your students are writing fairly long essays in the target language, but there are lots of mistakes in them, mistakes they often can self-correct when you point them out to them. A lot of mistakes are recurrent ones; they relate to things that you have ‘taught’ them over and over again in lessons and through your feedback and you have spent a lot of time drilling in; things like forgetting to make adjectives and nouns agree; wrong conjugations of verbs; omission or overuse of the definite article; wrong word order; omission of plural endings.

You are frustrated because you feel they should master them by now.

Other mistakes refer to function words, such as prepositions and conjunctions, things you may have not have emphasized enough and in true earnest are not massively important – but still annoy you.

So what is the way forward?

Before attempting to answer this question, let me remind you of the way the brain of an average student who has been taught grammar explicitly and vocabulary through single words processes language as s/he writes:

1- The proposition is created

2- The words that convey that idea are activated and retrieved from Long Term Memory

3- The words are then temporarily held in Working Memory for processing

4- Whilst in Working Memory the words are assembled together in the correct syntactic order and grammar rules are applied

5- The product of the previous phase is physically translated into graphemes

The above described process is quite cumbersome when it doesn’t occur automatically and the brain has to use every little bit of Working Memory’s very limited resources in order to execute each and every process involved.

Let us remember that WM can only process four items at any one time. If the students are used to handling single words – not chunks –  in their language production, that means 1 item = 1 word. This means that if the sentence the student is holding in WM in phase 4 is longer than four words the brain will have issues in monitoring the accuracy of the whole sentence.

For instance: take a French intermediate student having to translate the sentence ‘Yesterday she did not go out with her friends’. Whereas an expert speaker will not have to consciously apply any of the many grammar rules that underlie the production of this sentence, the intermediate learner will have to devote conscious attention to each and every step and take many micro-decision, i.e. Which tense? Is this a verb requiring the auxiliary ETRE or AVOIR? Where does the negative particle ‘pas’ go?  Etc. all these decisions will compromise the speed of execution; and to make things worse, due to the limitations of WM’s span, the items towards the end of the sentence are likely to end up receiving less attention and monitoring and therefore to be more vulnerable to error. Add to this that words decay from Working memory in a few seconds…

Another set of items that is more likely to receive less attention and hence will be more vulnerable to error will be those linguistic features whose contribution to the understanding of the sentence is less crucial; such items include function words, e.g. determiners, conjunctions, prepositions, etc. but also the execution of adjectival agreement, pluralization of nouns and verb conjugations. In other words, the brain will only focus on what is crucial for the expression of meaning – a sort of survival mechanism.

The whole issue is exacerbated by the fact that in the case of English learners of French, Spanish and other highly inflected languages, there is little transfer from the first to the target language in terms of the micro-skills involved in the execution and monitoring of agreement and verb conjugations – as in English you do not make nouns and adjective agree in gender and verbs are not highly inflected (i.e. conjugation are much simpler).

What is the solution then?

In answering this question the starting point will inevitably be another question: are your students actually ready to write extensively , considering the processing limitations just discussed?  Have you ensured that your students:

  • have received masses of practice in producing language under time constraints?
  • have been taught to produce language in chunks rather than single words?
  • have routinized verb formation under time constraints in context? Not simply being able to produce verbs in isolation?
  • have ‘formal accuracy’ firmly in their focal attention as they speak and write? – this is very important if you aim at high levels of formal accuracy. Bear in mind that they will only pay attention to form if you make a big issue out of it in every aspect of your teaching, from pronunciation to spelling, from word endings to sentence order. Way to often, the only time teachers really focus on accuracy is in their corrective feedback – that is a serious shortcoming.
  • have automatized at least the most basic forms of agreement?
  • can spell? – a students who is not confident with spelling, will have to focus attentional capacity on this level of production, which will eat into their WM’s processing capacity
  • have processed the language (grammar structures and lexis) that you expect them to produce in their essays extensively, through masses of exposure to comprehensible (95 % accessible without dictionary) input and productive practice under time constraints?

If the above pre-requisites have not been fulfilled, it is highly likely that your students are simply not ready to write that long essay – at least not as accurately as you expect them to. Chances are they are already doing a good enough job as it is considering their level of proficiency and the training in essay-writing they have received.

After all, they can only put down on paper what you taught them, right?

Hoping that your corrective feedback is going to do the trick is naïve to say the least. Yes, it will increase their awareness of what mistakes they make and maybe will sensitize them to the issue of accuracy; but only very few of your students will massively improve as a result of your corrections or reflection on corrections.

Correction may only work when it becomes remediation, i.e. sustained long-term instruction which targets the problematic items through extensive exposure and skill – not knowledge- based practice.

Concluding remarks

In my next post I will deal with the strategies that you may want to implement to tackle the deficits in your students’ writing. In the meantime, do consider the question in this post’s title: Are they really truly ready to write that essay or am I pushing  them way beyond the boundaries of their processing – not necessarily knowledge – capacity. Is asking your students to produce fairly long essays at the stage they are at in their proficiency development a productive way to foster their development as writers? Is it likely to erode their self-efficacy and motivation?

You might reply that exams are getting closer, only a few months away and you have no choice. Well, in that case, you may have to change the way you teach your students and prepare them for that task.

To start with, ditch single words lists and teach high-frequency chunks; increase the focus on formal accuracy; practise reading and listening for modelling (e.g. through narrow-reading tasks) rather than quizzing purposes (i.e. provides tons of comprehensible input in the aural and written texts you give your students); provide tons of practice in agreement, conjugation and function words usage through micro-writing tasks; give them a lot of text reconstruction tasks (like the ones you can find on the great Textivate website).

More on this and other strategies in my next post on writing.







More on universals, desirables, controlled input and implicit learning

This very concise post was motivated by the fact that many colleagues seem to have misunderstood what my ‘universals’ are about and by the high level of interest for this strategy shown by the teachers who attended my recent rounds of workshops in England and Australia.

My ‘universals’

So let me spell it out clearly: the ‘universals’ are high surrender value grammar structures, lexical patterns and/or functions that you feel your students are currently not learning effectively due to insufficient exposure or practice in your extant input.

By making such language items your ‘universals’, you commit yourself to embed them systematically in your daily input, week in week out from the beginning to the end of the year. This entails embedding them in your classroom talk, in your learning management routines (from instructions to oral and written feedback), in your every resource, in your every receptive and productive task.

Hence, my ‘universals’ are not – as some seem to believe – the dream list of ‘recyclables’ (key connectives, high frequency verbs, useful idioms, etc.) cutting across all topics that many teachers put on a sheet for their students to refer to whenever they write an essay or deliver a presentation. Far from it.

Rather, they are a very effective strategy to provide extensive exposure to and practice with language items that you are not currently teaching successfully; they are not to stay on a list that your students keep in their books ; you MUST ensure they are part of the comprehensible input you feed your students every day, and of the output that you push out of them.

For example, in my context, I was dissatisfied with my pre-intermediate students’ mastery of the French negatives and phrases used to compare and contrast; with their ability to create questions and with their repertoire of verbs. So, in every single lesson of mine I now make sure that my students get tons of implicit exposure to these structures and patterns across all four skills. Here is the full list of my year 7 French universals (see figure 1 below).

Figure 1 – year 7 French Universals



The hoped-for outcome: that by processing receptively and productively these items for a few minutes every day the whole year through they will become so familiar and embedded in my students’ cognition that by the end of term 3 they will be acquired by most if not all of them.

My ‘desirables’

My ‘desirables’ are also language items that I aim to teach implicitly. However, differently from the ‘universals’ they are aimed at only the most talented, inquisitive and proactive of my students. They consist of structures or patterns that are more complex and advanced and which I include in my input on a daily basis too hoping for the top 5 – 10 % of my students to notice and pick up.

An example of ‘desirables’ for my year 11 students refers to the subjunctive and patterns such as ‘bien que / sans que / pourvu que + subjunctive’ which I have been planting in my resources since last year. Another example, with the same group: the pluperfect indicative.

Comprehensible controlled input and feasible pushed output as crucial to the effective teaching of universals and desirables

Obviously, you cannot implicitly teach universals and desirable without constantly recycling them through controlled input which is highly patterned, repetitive and 95 % comprehensible. Authentic texts will not include as many instances of your universal or desirable items as you need to drill them in effectively. Moreover, authentic texts rarely include comprehensible input.

By the same token, unstructured tasks will not ensure that your students will include them in their output; hence the need to provide extensive productive practice which elicits their deployment task after task, the easiest and safest way of achieving this being oral and written translation tasks involving feasible output (e.g. those discussed here).

In my next post ‘The two keys to effective language teaching and learning: controlled input and pushed output’ I will elaborate on this further.


This post was motivated by the fact that some colleagues have equated my universals to Barry Smith’s top tens or to examination boards lists of essential structures/lexical patterns, what teachers often refer to as ‘recyclables’.

Unlike the above, my ‘Universals’ and ‘Desirables’ are what YOU want them to be; I conceived them as a way to keep in my attentional focus the structures and/or lexical patterns I was teaching less successfully so that they would feature day in day out in my input and in my students output.

For the universals/desirable strategy to truly impact your students they need to be recycled systematically and methodically in your lessons. The more they occur in your input and your students’ output the better.

The universals ILRs (implicit learning routines) I use in my daily practice ( e.g. ‘grumpy time’ or ‘question time’) take very little time and allow me to never lose track of my universals, whatever the topic-at-hand is (see here for discussion of my ILRs). Scaffolds like the one below (see figure 2) will assist your students when they are asked to produce the universals orally or in writing.

Figure 2 – scaffold I use to model and encourage use of negatives


I have tested this strategy many times over and it has never disappointed me.