Skip to content

Contrastive analysis for error correction

How often have you listened to learners talking and thought to yourself: ‘you made that error because that’s the way you say it in your mother tongue’?

Here’s an example of addressing such errors from a recent one-to-one lesson, followed by some thoughts on contrastive analysis.

My student and I had been talking for some time. As I was making notes I heard a succession of errors that appeared to come from his L1 – Japanese.

When the conversation wound up we went back and analysed the language. My usual method of delayed correction is to write up errors and ask the students to suggest improvements. This time, I tried a ‘stealth’ method of error correction – you may not wish to do this with learners who are sensitive to error!

1. Write out corrected versions of the utterances
2. Show the learner(s), but don’t mention that they’ve been corrected.
3. ‘Here are some sentences from our conversation. Under each one, write how you would say it naturally in L1.’
4. At this point you have two correct versions of each sentence, L1 and L2. Next, underneath each one, write what they actually said.
5. Ask the learner(s) to go through, identify their error in each case and trace it back to how it would be said in their L1. Talk about the causes of the error.
6. Set a follow-up task to practice. Perhaps a writing consolidation for homework.

Here’s the page. 1=corrected English; 2=Japanese equivalent; 3=original utterance.


The third group is an interesting case. It turned out I’d misheard ‘although’ as ‘also’! This was principally a problem with word stress, but heightened by the fact that he pronounced ‘th’ as ‘s’. Conveniently, this is actually an L1 interference problem, this time of the more predictable phonological variety.

And the follow-up:


I’ve always been quite interested in this aspect of SLA because in my own language learning I tend to value rather highly an understanding of how something is expressed differently between L1 and L2. It helps me to make sense of language, understand its inner-workings.

In class, drawing comparisons between L1 and L2 has always seemed to me a logical thing to do. If a Japanese learner proclaims ‘I’m becoming happy’, and your knowledge of their L1 tells you that you often ‘〜になる (become+emotion)’ in Japanese, then would it not make sense to point this out? Things like this make me question the default position of many teachers of ‘no L1!’


Reviewing and scaffolding language with one-to-one learners

Here’s a simple activity for looking at and deepening understanding of a chunk of language. It works particularly well with one-to-one classes because it needs a fair amount of teacher input. You could use it to examine emerging language, or to review something from a previous lesson.

Take a chunk that you want to explore in more depth and write it out in it’s base form the centre of a blank piece of paper.

For example: from the sentence ‘I thought he would take it personally’ you might end up with ‘take it personally’ on your paper.

Teacher and student(s) take it in turns to add a word each to the chunk. It could be to extend it (‘he took it very personally’) or in the form of substitutions (‘take it badly’).

It’s important to talk about the language that comes up and talk about what isn’t possible just as much as what is. Stretch the learners to think of different contexts and applications for the language, scaffolding all the way. What you end up with is a rich exploration of language that can be recorded in lexical notebooks and used for homework – eg ‘write 10 possible sentences from the exercise’.

A great homework task would be to make a language plant.

Here’s how my above example ended up in the lesson:


The state of my one-to-one teaching, and a way forward

One-to-one teaching

In this post (originally written for the British Council Teaching English Facebook page:, I’m going to briefly describe how I see Dogme situated in common practices in one-to-one teaching, describe a lesson that I taught last weekend, then finally discuss how my own teaching has developed recently,  perhaps influenced by my experiments with Dogme. I started writing my blog as an alternative to a personal teaching journal, a way to reflect on my lessons and my teaching. It seems that people really enjoy reading other teachers’ reflections, though, so I’ve started writing about my teaching in general. I hope you find it interesting and please feel free to comment – blogs thrive on the interaction between people out there!

Many people’s reaction upon encountering Dogme for the first time is: “Well, I pretty much do that anyway.” They’re usually referring to their private teaching. I’m sure most of us can relate to an approach to one-to-one lessons where we talk with the learner, scaffolding their language, essentially helping them say what they want to say while talking with them as much as possible, perhaps stimulated by an article or similar that the learner has brought to the lesson. We might then come up with a few exercises on the spot based on what they’d had problems with and then set something for homework that would allow them to practise further.

If your own private teaching is anything like the description above then you already have Dogme in your teaching blood!

1) your lessons are driven by conversation on relevant topics

2) you use your learner’s emerging language as a basis for language work

3) any materials you use are relevant to the learner, either because they’ve brought it along to the lesson themselves, or because you’ve used your familiarity with them to select a text suited to their needs or interests.

This method of teaching deals with most General English students’ learning needs very well, and I think a diligent teacher can do well with this. This is true only insofar as the learner is getting personal attention. Transplant this into group teaching, and you have a very different kettle of fish because it’s impossible to attend to everyone’s needs, as they emerge, all of the time. A lot of the approaches to navigating Dogme lessons that I write about in my blog are basically ways to stimulate conversation democratically – ensuring that all learners are engaged, contributing and hopefully benefitting more or less equally. Seemingly innocuous activities like: “tell each other 3 good things and 3 bad things about your weekend” actually fulfill an important function, because you’re able to immediately move the focus of the group away from you, the teacher waiting to ‘teach’, and towards the learners themselves and what they have to offer to each other.

But this post intends to deal with one-to-one teaching.

After an unreasonably long time between lessons, I saw one of my long-term students Naoko last weekend. We were rather short on time (1 hour), and I set about the lesson in my usual fashion. It was only when we’d finished and gone our separate ways that I reflected on the lesson and realized how different it was to how I would have taught a similar lesson even 6 months ago.

It’s interesting for me to note that, despite what I’ve written above about an ‘organic’ approach to teaching one-to-one lessons, the more experienced I become, the less ‘free’ my lessons are becoming. In fact, it’s clear that I’m imposing quite structured series of activities on the learner during the lesson. Why the move in this direction? On the face of it it seems rather anti Dogme principles. In group teaching I work so hard to create communicative situations, but with an amenable one-to-one student it’s much easier; the information gap between teacher and learner (L2 vs L1) is there from the start, thereby generating the authentic communicative situation – the need for the learner to make herself understood. So perhaps I feel that, with this burden lifted, we can get down to the nitty gritty language work more quickly.

The lesson

Level: Advanced. Time: 60 mins

Pre-lesson: there is no preparation. However, I have a photocopy of her entire lexical notebook which I look at before the lesson, noting down a selection of words and phrases that I know will be useful for her to revise and that are likely to come up in conversation. I’ll often choose lexis with some kind of function in discourse e.g. ‘at the same time’, ‘all in all’.

1) ‘So what’s been going on since the last time I saw you?’

2) She tells me about her father not being so well and a trivial family feud

3) I don’t interfere linguistically. I listen and respond as naturally as possible, asking a few questions.

4) We her situation and possible solutions to the problem, throwing in some phrases and expressions that I think she’ll like, and fitting in some language for recycling from previous lessons . Some discussion of those follow.

5) Among other things, I give her some ‘sentence frames’ based on what she was trying to say: ‘I don’t want to antagonize him, but I don’t want to avoid the issue either.’ This becomes: ‘I don’t _________, but I don’t ________either.’ We take a detour and come up with some substitutions for this frame.

6) ‘Right, tell me it all one more time.’

7) Naoko repeats the story. She’s a little reluctant at first, but appreciates the value in it. If she forgets bits I ask for details in as natural a way as possible – clearly there’s an element of ‘display’ here, but it serves a purpose

8 ) Some more feedback and discussion. ‘Ok, and one last time.’

9) It’s very good this time. She’s very bright and is already incorporating most of the language we’ve discussed, making very few mistakes. What to do? Really the only thing that’s setting her apart from a native speaker at this point (other than her accent) is that it’s sounding too good. It sounds prepared, like the Best Man’s speech at a wedding. “I first met Peter…”. Furthermore, in the back of my mind I’m thinking:  ‘as good as this is, she’ll have forgotten most of it by next week.’ I decide she needs to relate what she’s saying to authentic speech.

10) ‘Now I’m going to tell you the story myself. Just listen and tell me what differences you can see between our two versions.’

11) I tell the story, making a point of telling it just as if I were in the pub, throwing in all the elements of natural speaking you’d expect: false starts, hesitation, repetition etc.

12) We talk about these features a little.

13) I used her iPhone to record myself telling the story. Homework is to transcribe it word for word, false starts and all!

Commentary – What am I doing differently?

The main thing: we really did much less than usual. There was very little superfluous chat, or any speaking that didn’t have a pedagogical purpose in the back of my mind. I’d usually spend some time just chatting with the student, warming into the lesson (who doesn’t?) but it’s very interesting how a bit of time pressure made me focus much more quickly and spot potential directions for the lesson. It just so happened that she had a story to tell me, but a simple ‘how was your weekend?’ could have brought about similar results.

I’ve been struggling recently with ideas for consistently recycling language from previous lessons. How can it be done naturally in Dogme? I don’t know how useful the above method is, or just how far it can work. Clearly some language is easier than others to slot into new contexts. Honestly, though, given that I have been guilty of not recycling language so thoughtfully in my teaching in the past, right now I’m just happy to be thinking about it!

The language I’m giving her is different to before. FIrstly, I noticed that I chose to focus on a relatively small amount of new lexical items over the course of this lesson. Perhaps 6-8 at most. There were so many things I could have brought up, but I resisted the temptation. This meant that, in the detours we did take to discuss language, we could go into sufficient depth.

The more experienced I become, the ‘larger’ the language that I focus on. That’s not a particularly eloquent description, but here’s what I mean. I wouldn’t have spotted/conceptualised language in a way such as ‘I don’t _________, but I don’t ________either,’ in the past. These ‘sentence frames’ (also known by a variety of terms) are an approach I’ve used a lot recently as they seem to be very easy for learners to get their head around. I think they’re open to criticism on some levels, for example many elements of the frames themselves can be changed (e.g. ‘she didn’t ____, but she didn’t ____ either). However, by focusing on language on this more sentential level you’re encouraging learners to think more holistically, which will ultimately impact on fluency.

I got her to repeat the speaking ‘task’ twice. This approach, borrowed from Task Based Language Teaching (TBLT), has many benefits and by the third time learners are usually really getting a handle on what they’re doing. Although I have always liked repeating tasks, I’m finding myself actually insisting on it now; not just once, though – do it twice! I have found that not all learners see the value in this approach initially, but it’s really worth sticking at it because when they come round they really appreciate it.

Dogme principles state that any listening work done in class should use recordings made by the learners themselves. This has seemed rather counter-intuitive to me, because what of the value in exploiting authentic texts? However, with this method of recording a ‘better other’ version of Naoko’s own language, I believe I might have found a cunning compromise. After all, it’s highly relevant to the learner (she came up with it!) and ‘meaning’ is already taken care of. She’s free to study the gap between her version and my version. By coming at the end of the lesson, I see another parallel with TBLT, in which the language focus stage of a lesson may come right at the end.  Being set for homework it provides a sense of continuation between lessons. Learning takes place in a continuum and does not end when the hour’s up. She’ll take this recording home and really study it. I have no idea what she’ll take from it linguistically, but there’s a rich source of language there that we can discuss in the next lesson and possible the one after.

Some questions to think about

How could this be continued into the next lesson?

What would you choose to focus on?

Would any of these ideas work in your own teaching?

Would you change anything? Why?

How do you recycle language from previous lessons?

Lesson 10 – Taboos

Today’s lesson really could have been straight out of a coursebook. In fact it probably was originally, in which case I hope this lesson serves as an example of how a coursebook idea can be adapted and exploited to make more use of learners as a resource.

The nature of the lesson emerged naturally but I think I was influenced by something I saw once in a fantastic resource book called Taboos and Issues.

The group came in and started talking about some health issues, and it emerged that they had just had their annual checkup arranged via the company. One student in particular was saying how he had to go back for a colon-related examination and started giving some details about samples that I didn’t really want to know about. He was slightly struggling to explain it all though and I thought whether I should fly with this as the lesson topic…… but I quickly rejected the idea!

I commented that it’s quite unusual in the UK to talk about health issues like this, to which they responded with some surprise that it’s quite normal in Japan.

The topic of taboos had presented itself on a plate. From there I went through the following sequence of activities which is hopefully exemplified in the board photo:

– what topics are delicate in Japanese society?

– Ss work together to rank them in order of sensitivity (ranking activities are great because they force the Ss to probe deeper and articulate their feelings, throwing up useful language in the process)

– Brainstorm some ‘inappropriate questions’ on each of the topics (lots of fun). I let them do this in L1 initially because they were finding it hard to conjure up examples from memory in L2. Interesting how language and memory/culture interact here.

– correct and drill questions

– now imagine situations in which these questions could legitimately be heard. This yielded some really interesting conversation as they explored the boundaries of their preconceptions.

60 mins

Commentary: Recently, I’ve been trying to find ‘language opportunities’ on which to pursue my Dogme lessons, but I took a different route here and followed a ‘topic opportunity’. In this case, although there was language arising during the lesson, it was the logical progression of the topic that carried it through, reminiscent of a Task Based Learning procedure. I once read an article on Dogme (or perhaps Teaching Unplugged itself?) which commented that Dogme and TBL share many parallels, and I experienced that first hand today.

Often, my Dogme lessons that don’t work as well as the ones I write about here fail to take off because I can’t come up with activities that are stimulating/logical/relevant enough off the top of my head, in the heat of the lesson.

Recently this has been a persistent problem and I think it’s because I don’t have enough good activity options ‘at the point of retrieval’. I’m thinking that a period of teaching carefully thought-out TBL lessons with well sequenced, progressive and orderly tasks would have very positive backwash to my unplugged teaching.

How have other teachers upped their game in this respect? It’d be great to share some teacher development stories. Journals? Activity banks?


Lesson 9 – Business and a nod to M. Lewis.

In the early stages of unplugged classes I usually find myself actively looking for the emerging topic or communicative situation that will provide the impetus for the rest of class. In today’s business class that didn’t happen because one of the Ss had come in with an email that she needed help understanding. She then went on to try to explain what it was all about, largely out of frustration and certainly not because she thought it would be an interesting lesson topic. It was a horribly complex situation regarding new import/export rules imposed by the Korean government and the chaos of a stranded shipment. Certainly not a lesson topic.

But as is often the case with these classes, as we know, beginnings and endings can be messy. People often don’t leave their baggage behind at their desk. They bring their working lives and problems right into the room with them.

The great benefit that Dogme has brought to my teaching is the ability to engage with these situations and turn them into an interesting and meaningful lesson for everyone in the class, even if the problem originated with one person. Not all that long ago, given the same situation, I would have continued the conversation for a few minutes, shown a bit of sympathy, and moved on to the prepared lesson (which had to be curtailed because of the time spent talking at the beginning).

So given a rather dull topic like this, what can you do with it? Tomoko was clearly frustrated with the problem and was really trying to explain the background.

1. I continued the conversation, involving everyone else, scaffolding to help get meaning across, taking notes all the time.

2. This continued for some time. I felt it was important to deal with the topic satisfactorily. We had to be on the same page or it ceases to be a genuine communicative exchange.

3. There was a lot of language that came out, too much to deal with. At that point I wasn’t sure where to take it.

4. Some errors on the board, Ss discussed and corrected while I reflected on the language myself. (Do you ever find that the appropriate language to work on eventually presents itself on its own?)

5. A lot of discussion over ‘they refused to’ and questions emerged about ‘deny’ and whether it was the same.

6. Quite a few phrases and collocations emerging. Board was erased, books closed, and a collocation matching exercise written up on the board. (see board photo)

7. Continued discussion over ‘refuse’ and ‘deny’. In the original story, the problems centred around the Korean government refusing to issue import papers. I decided to go with this because of its relevance to the story and Ss emerging interest in the language.

8. An example sentence of each written up, patterns and substitutions elicited.

9. Up to this point, fairly standard lexical work. But this language isn’t easy for them, so how to deepen it? Variation of an idea from Lewis (1997) – ‘what you think/what you say’. Having already covered the reported language (“they denied us more time”), Ss discuss what was originally said by the speaker (I.e. “you can’t have more time”).

10. Contrastive analysis. Ss translate the examples into Japanese and discuss nuances. Not as straightforward as I’d expected. Turns out you can say ‘refuse’ in three different ways in Japanese depending on degree of politeness.

I feel like we’re getting there.

11. Ss test each other on the language. One person reads the Japanese sentence and the other has to say it in English. Ss correct each other and I drill each correct sentence as it is produced. This is great fun.

60 mins

Comments: In reflecting on my own progress in Japanese, I’ve often been surprised at how hard it is to cover even a small amount of language thoroughly in the space of an hour. There has never been any sense in my own learning trying to learn too many new items in one lesson, after all you can do that on your own.

What the teacher can bring to the lesson is strategy to deepen your knowledge of the language. Was it Willis who said something like “little language, lots of talking/using”, something like that?

Given the quantity of labguage that emerges in an unplugged lesson, Dogme teachers should be sure to ask themselves if they are dealing with language in enough depth. Are we reducing uncertainty of language enough for learners to have a clear understanding?

Translation reveals a lot that we may not be aware of as teachers. People translate in their heads whether you like it or not, so as teachers we should explore what it means for learners. In this case, I discoverer that the complexity of ‘refuse’ in Japanese meant that Ss were likely to need more work on meaning.

The concepts of a Lexical approach are very compatible with Dogme. Of course we already knew this, but just flicking through the exercises in Lewis (1997) reminded me just how good a resource it is. Dogme teachers would do well to keep a copy close to hand out of lessons… Or even in lessons.


Lesson 8 – YL Dogme

I’ve seen a number of posts recently asking about Dogme with YLs. The immediate problem that YL teachers face is that classes cannot usually be ‘conversation driven’, meaning that inspiration has to be found from elsewhere.

How can you do this with, while trying to stay ‘materials light’?

This is a lesson that I did with a group of 40 junior high school kids aged 11-12. It’s based on a lesson designed by a colleague, which originally included a a language presentation element.

However, I felt that it could all be done ‘unplugged’, based solely on the experiences of the people in the room.

The earthquake and tsunami in Japan earlier this year had a tremendous effect on the national psyche, but it has been especially traumatic for kids, who now have to perform regular earthquake drills in school. It’s a very real element of their lives and the perfect basis, I thought, for a Dogme lesson!

Lesson 1
– Elicit the idea of earthquakes
– Remind Ss of their last drill and ask what they’d learnt to do in an earthquake
– Set them brainstorming ideas in groups
– Feedback to board, translating/reformulating their ideas into do/don’t statements (e.g. Don’t go back inside, go to a high place)
– Drill
– Ss copy language into notebooks
– Clear the board, then give them a dictation exercise (or similar) to reinforce language, which is quite challenging
– tell Ss that in next lesson they’re going to make an earthquake advisory poster which will be displayed in the school
– Ss work in pairs to sketch some sample ideas for posters

Lesson 2
– revise language from last lesson
– Ss choose their favourite idea from last lesson and make their poster

Even the weaker Ss were involved and engaged in this lesson and although the conversation took place largely in L1, the English that emerged was focused and meaningful for them. Perhaps this is why language from this lesson appeared to be especially well retained in subsequent weeks.

Here’s a sample of the work they produced. I find their imagination inspiring and their artwork must have created in them some very strong positive associations with English (one of the most challenging things to do with high school classes).

This slideshow requires JavaScript.







Lesson 7 – A black comedy (pure Dogme?)

Tomoko comes in first and starts telling me about a trip to the theatre at the weekend and a rather black comedy that she saw.

It was tough for her to explain and the challenge clearly frustrated her. I listened as she told me about the story, scaffolding her language, making notes and involving the other Ss as they arrived, inviting them to ask questions to catch up on what they missed.

By the time she had finished, she looked like she wanted to forget the whole thing – she really struggled to get to the end of her story. The other Ss obligingly told her that ‘even in Japanese this would be difficult to explain!!’

It was a clear learning opportunity and I set about reconstructing the story in mind-map format on the board, re-eliciting words and phrases she had wanted to use, or that I had fed in.

They were gripped. I think Tomoko felt a little guilty/embarrassed that the lesson was developing around her story!

The board photo tells the story. Word classes and patterns were elicited, some drilling and phonology work, and a few other bits and pieces. This all took 15 mins or so.


When we finished, Ss took it in turns to retell the story to the group (only 3) using the language from the board, while others listened/monitored and occasionally corrected mistakes. I didn’t interfere much.

The board was then wiped clean, books closed, and the story-telling repeated, one by one. Their performance was astonishing. They all retold the story almost perfectly, recalling all the lexical items and phrases (of which there were at least 15). And this from a group that on the whole really struggles to retain language.

With more time I would have done more work on expression, but I was already thrilled with such a good outcome after 60 mins.

Ss were delighted and rather shocked at what they’d achieved.

The lesson felt very rounded and balanced in terms of Iearner input, language work and practice, and I’m growing in confidence as I become more experienced in deciding the direction of the lesson from minute to minute and choosing the most appropriate activities.