Why do our L1-English learners of French/Spanish find it hard to acquire agreement rules? What can teachers do to facilitate the process?


In a previous post I already dealt with the dichotomy declarative knowledge vs procedural knowledge and control. To put in a nutshell, the former refers to knowing the set of ‘rules’ governing the use of a given target grammar or lexical structure (having its mental representation) whereas the latter refers to its effective application during real operation conditions (e.g. in spontaneous speech or writing under time constraints). As I have often reiterated in many of my posts, language learning ought to aim at bringing about high levels of target language control (as close as possible to automatization), whilst viewing declarative knowledge as the necessary starting point in an L2-learner’s journey towards acquisition.

In my experience, students find it relatively easy to grasp the rules underlying the application of noun (or pronoun)-adjective rules but rarely manage to acquire effective control, and even at A-level and University many mistakes continue slipping into performance (especially in oral output). Why is it? And what can be done about it?

In a previous post I discussed how agreement errors are often due to the fact that, in less expert L2 speakers/writers, whenever Working Memory experiences cognitive overload, the brain tends to focus only on the most semantically salient features of the output (the ones that convey most of the intended meaning) and neglects the features which do not contribute much to meaning. However, this is not the whole truth. The picture is much more complicated than that.

Let us look at the cognitive operations and knowledge involved in the process of applying adjectival agreement rules in the production of L2 French/Italian/Spanish/German output (i.e. speaking or writing) under real operating conditions (henceforth ROC). The L2 speaker/writer must:

  1. Retrieve the required French adjective;
  2. Remember to make it agree with the noun in terms of gender and number – which is not always straightforward as they may be relatively far from each other – separated by a copula and an intensifier, for instance;
  3. Know whether the nouns is masculine or feminine;
  4. Know whether it is irregular or regular;
  5. Apply the rule;
  6. (in speaking) pronounce it correctly / (in writing) spell it correctly.

These are quite a lot of cognitive operations to perform under ROC. To top it off, the cognitive load posed by these operations is exacerbated by the fact that there are often other permutations that the learner will have to execute in the same sentence (e.g. subject-to-verb agreement).

As pointed out in previous posts, if our learners keep making this kind of mistakes day in day out whenever they engage in spontaneous or pseudo-spontaneous communication, the errors end up being fossilized (automatized) and incorporated permanently in their Interlanguage. This explains why a lot of L2 learners keep making those mistakes all the way up to university. So what can be done to fix this problem earlier on? Lots of old-fashioned drills? Or how about, as the ‘Krashenites’ amongst us would suggest, exposing the students to lots of comprehensible input and avoid involving them in any language production until later stages in the instruction process? Neither of these solutions are in my opinion, a bad idea. In fact, any sound approach to this issue, would have to involve a bit of both.

To come up with an effective solution one should, in my opinion, consider first the three main psycholinguistics causes of the issue, which refer to the six operations listed above.

  1. The gender of nouns – the notion that words can be masculine and feminine (or neuter if one is learning German) is completely alien to an L1 English native speaker. Yet, how often and how strongly is this notion firmly placed in the students’ focal awareness and ‘drummed in’ across all four language skills during the early stages of acquisition – and later on, too – by MFL instructors? Enhancing their focus on this notion does not simply involve teaching them the gender of the target nouns; it also involves changing their mindset, the way their cognition works. Hence, teachers must ensure, since the very early days of L2 learning, that students are constantly reminded of this concept, both explicitly (e.g through work on noun morphology) and implicitly (e.g. through colour coding).
  1. Focus on word endings – the anglo-saxon brain is wired to focus on the beginning of words; hence, instinctively, an English native speaker would focus his/her attention on the opposite end to where s/he should indeed be focusing it on. This also entails another disadvantage: students may not learn much from any L2 written input they read since, by focusing mainly on the beginnings of words, may not notice the endings in the texts at all.To get an anglo-saxon brain to invert the ‘instinctive’ focus of its attention is no easy task, especially with adult learners. This process will require extensive day-in-day-out scaffolding and practice.
  1. The saliency of agreement – this issue compounds the problem identified in point 1, in that an L1 English speaker is not only at odds with the notion of gender, but will also find the notion of agreement unfamiliar and redundant. Hence their brain will automatically place the saliency of agreement low down in their list of attentional priorities. The challenge for teachers is to ensure that agreement is constantly in the learners’ focal awareness until it becomes ‘second nature’ – as it is for any French, Spanish, Italian and German native speaker. By making the application of agreement become ‘second nature’, I mean that whenever an adjective is retrieved by Working Memory, a ‘program’(or Production, as it is called by Skill theorists) in the learner’s brain is automatically activated  that operates something like this:


        If condition: if I use an adjective in a phrase/sentence…

       Then condition: …then I must make it agree with the noun it modifies


The speed at which the brain will activate the above Production (i.e. the extent of its Proceduralisation) will play a big role in determining how efficiently and effectively the agreement rule will be applied.

The implications for teaching are pretty obvious. MFL teachers must focus on developing processing efficiency under ROC (i.e. cognitive control), whilst addressing the three issues just discussed, by moving them into their learners’ focal awareness until, after day-in-day-out scaffolding (in the way of reminders) and practice (only a few minutes a day), they become automatic. I have already discussed fairly extensively how control can be enhanced in a previous post (“Control – the most neglected, yet most important factor in MFL grammar teaching”); as far as the other issues are concerned, here are a few possible teacher tactics. The reader should bear in mind that in my approach one should always start with receptive skills and move on to the productive ones at a later stage.

Focus on gender – here are some suggestions on how to focus students on gender:

  • Present masculine and feminine nouns always with the (indefinite/definite) article or any other determiner (e.g. mon/ma) and using different colour coding (this is common practice in many MFL classrooms);
  • When providing vocabulary lists, make sure that the masculine and feminine nouns are grouped separately (you may use colour-coding as background to enhance the contrast) ;
  • Model and practice extensively ‘rule of thumbs’ which may work as ‘aide-memoire’ in the identification of the gender of nouns (e.g. noun endings in ‘ion’ are usually feminine). Engaging inductive activities can be staged in class whereby the students are given lists of words and with the help of dictionaries need to work out by themselves such rules of thumbs.
  • After involving the students in a reading or listening-based activity, get them to identify (based on their determiners) the gender of a set list of nouns whose gender you want to focus on;
  • Involve the students, on a regular basis – I do one every single day – in quizzes based on gender identification e.g. odd one out’s (given three nouns, spot the feminine one) and gap-fills (where the article must be inserted);
  • Give the students a short passage containing X number of mistakes with (the gender of) articles or other determiners and challenge them to find them under time conditions with the help of the dictionary. This can be done as a way to practise the modelling of ‘rule of thumbs’. Students usually enjoy this activity;
  • The classroom environment can be used as a way to remind the students of the issue and to display any rule of thumbs modelled or worked out by the students.

              Focus on adjectival endings

  • Colour-code feminine endings – as well masculine endings when dealing with irregular adjectives (e.g. travailleur vs travailleuse);
  • When providing vocabulary lists include both the feminine and masculine endings of the target adjectives(it is tedious and time consuming but it pays off);
  • Listening activities involving focus on endings should be carried out regularly (e.g. minimal pairs, where the feminine and masculine forms of the same adjective are contrasted);
  • ‘Error hunt’ tasks where students need to identify a set number of agreement errors in a text – students usually enjoy this kind of activities;
  • Old-fashioned drills (e.g. multiple choice gap-fills; ending manipulation tasks; translations, etc.)
  • Give the students a checklist with the following guiding questions, for example, as a way to scaffold the focus on adjectival endings and agreement when they are producing written output, e.g.: (1) Which noun does the adjective refer to? (2) Is the noun feminine or masculine, singular or plural? How do you know? Have you double-checked, if in doubt? (3) Is the adjective regular or irregular? Have you double-checked, if in doubt?

Placing ‘agreement’ in the students’ focal awareness – As far as this issue is concerned, the above activities, if practised regularly, ‘should do the trick’. The above mentioned activity involving work on ‘authentic’ essays written by previous cohorts of students containing numerous mistakes with adjectival agreement could be used as a reminder of how common this type of error is. Also, in setting targets as part of feedback on writing or speaking, one of the three or four targets identified should include adjectival agreement if it is a recurrent source of error in the output. A narrow-focus corrective approach, whereby the feedback and feed-forward on a student’s written output centers mainly on one or two issues only (see my article on this approach, could be implemented, too. In this approach, the students could be focused solely on agreement issues for a few weeks so as to channel all of his/her attentional resources in the editing process only on this aspect of grammar accuracy.

In conclusion, the cognitive challenges posed by the acquisition and application of agreement rules are manifold. In this article I have endeavoured to outline a few. Most teachers do address such challenges, in my experiences, but not consistently and extensively enough to prevent them from causing these errors to recur and become fossilized in their learners’ interlanguage. Some practitioners adopting strong CLT approaches may not feel that agreement errors are important enough to deserve the allocation of dedicated teaching time in each lesson. I can relate to this argument, as I agree that fluency should come before accuracy as a priority.

However, agreement mistakes, when they are recurrent, can be stigmatizing and irritating to native speaker readers or listeners and may be interpreted by them as signs of poor linguistic competence. Hence, I advocate that a few minutes’ work on the above issues should feature regularly at the early stages of L2 instruction until one feels that the learners have finally acquired a sufficiently high level of focal awareness of and control over this structure.

Cognitive control – the most important, yet most neglected factor in foreign language grammar teaching

download (2)

Cognitive control, as I mentioned in previous posts, refers to the ability to perform a task in real operating conditions. For instance, an L2 learner of French who has had only a few lessons on the Passé Composé may be able to conjugate the verb ‘Aller’ perfectly in the context of a gap-fill exercise; however, when required to use it in spontaneous speech he will produce wrong utterances like ‘J’allé’ or ‘J’ai allé’. Most experienced teachers will be very familiar with this phenomenon – and with this specific mistake.

The reason for the mistake is that the learner has not yet acquired cognitive control over that specific form of the Passé Composé; he has a clear idea of how to form it, has the correct mental representation of the rule (or Declarative knowledge, as Skill theorists call it), but cannot apply it in real time when his brain has to juggle all of the following demands in the very short time available to him to ‘stay’ in the conversation:

  1. Understand what the person he is talking to is saying;
  2. Plan what to reply;
  3. ‘Fetch’ from Long-term Memory the French words that match that plan;
  4. Arrange the words into a syntactically correct sentence;
  5. Store that sentence in Working Memory;
  6. Modify word endings when necessary (i.e. verbs must be conjugated, feminine/plural endings added, etc.),
  7. Evaluate its accuracy, which means monitoring the sentence whilst rehearsing it in Working Memory;
  8. Pronounce it correctly.

This is a tall order for a novice foreign language speaker as it requires the ability to orchestrate many skills at the same time. Teachers often take it for granted as they are, after all, very good linguists and may not remember how they themselves struggled with that as learners…

To go back to the example I gave at the very beginning of this article, what is ironic is that what the student who gets the ‘je suis allé’ wrong in spontaneous speech will often get by his teacher is a reminder of the rule – that the boy already knows – and more gap-fill exercises – which will not prepare him for effective performance in spontaneous speech at all.

But how do we develop L2 learners’ high levels of executive control (also called ‘Procedural Knowledge’) over a given structure or skill? Surely not simply by getting them to memorize scores of conjugation tables and practice through gap-fill exercises. These activities can be useful as a starting point along the acquisition continuum, but ultimately, the only way to acquire the ability to apply a grammar rule in real operating conditions (i.e. spontaneous speech; essay writing under timed conditions; chatting on the web, etc.) is practice which starts with very easy recognition tasks and culminates several lessons later into more challenging unplanned communicative tasks (e.g. spontaneous conversation). This is an example of a sequence of (easy-to-prepare) activities to develop control in speaking for a year 9-10 class, which would take at least three lessons:

  • Presentation / Modelling of the target rule: Explicit (e.g. typical Power Point presentation explaining the rule) or Inductive (e.g. students are given a text or sentences with examples of the target rule and they must work it out by themselves)
  • Lots of receptive practice where students see examples of the target grammar structure’s application in the context of a written passage;
  • Gap-fill exercises and/or audiolingual-style drills (e.g. students are simply required to repeat a set sentences but change a verb or adjectival ending)
  • Easy (English to French) oral translations (e.g. students are given very basic role-plays in English to put into French orally);
  • Picture-based tasks (i.e. given a very clear and simple picture, students have to briefly describe it using the target structure);
  • Structured ‘narrow’ tasks eliciting basic responses containing the target structure (e.g. basic conversations, surveys or ‘Find someone who’ tasks including questions such as ‘Qu’est-ce que tu as fait hier soir au cinema?’);
  • Challenging GCSE style role-plays with cues eliciting the target structure;
  • Typical unstructured communicative tasks eliciting spontaneous speech (e.g. GCSE style conversation tasks)

As the above sequence clearly show, the implication is that the acquisition of high levels of executive control over any grammar structure requires a lot of practice through a model that goes from ‘easy’ receptive tasks, to gradually more challenging production tasks which are fairly easy and highly structured to start with and become increasingly demanding on Working Memory capacity. In my experience, learners are way too often required to skip steps and go much too soon from the gap-fill / easy permutation stage to fairly challenging GCSE conversation-like tasks. This is very much the equivalent, in real life, to showing a person how to drive in a motionless car and ask them a few minutes later to drive the same car flat out on a trafficked highway.

In conclusion, the notion of control that I have dicussed in this article should be heeded by language instructors to a much greater extent than it is currently done in many L2 classrooms. One would not claim to be able to play tennis after reading a book about it, or after having bounced the ball with one’s racket against a wall a few times. However, this is pretty much what many teachers do when they limit their teaching of grammar to a presentation of the rule, a few gap-fills and one or two rounds of questions around the class to check understanding – followed  maybe by a short essay or narrative eliciting the use of the target structure.

The achievement of high levels of procedural knowledge must be the ultimate goal of foreign language learning; hence, the development of effective cognitive control must be at the heart of everything we do in every lesson.

Eight must do’s of MFL project-based learning

images (3)

Those who have read my previous blogs will know that my espoused methodological approach to MFL instruction is situated in the Skill-building paradigm and is a combination of CLT and Focus on forms (not ‘form’). However, I have indeed tried out Project Based Learning on a number of occasions in the past, with varying success. Although I see serious advantages to this pedagogical approach there are serious threats to its effectiveness, which are partly intrinsic to the nature of the tasks set and partly to the way they are implemented by the teachers and managed by the learners.

The following are, in my opinion, the 8 most important steps to take in the implementation of Project Base Learning in order to control for such threats and enhance its learning impact on L2 learner proficiency. In what follows, I will only focus on pedagogical recommendations which relate to the specifics of target language acquisition I will not concern myself with other aspects of the planning of PBL, which pertain to the realm of metacognition, collaboration, enquiry and other generic skills.

1. Make sure that you decide on a core set of target language items and plan carefully for their recycling

One of the dangers of project-based learning (henceforth PBL) is the relative lack of control the teacher has over the language the students will process receptively and produce. Whilst one might see it as an advantage, as the students are being creative with the language, there are two problems with this. Firstly, a learner needs to process any given lexical items several times (somewhere between five and ten times) in order to learn it. Secondly, if we are to test the learners fairly at some point to assess how much has been learnt in the process, the students must have been exposed to a common core of vocabulary and grammar structures.

Insufficient recycling of language is the most common and serious pitfall of PBL. Students will not learn much. I used to have a colleague long ago whose students became very creative indeed in terms of graphics, photography, filming and use of digital media. However, their retention of anything they had written in the process was very poor. Had my colleague found ways to recycle at least a core set of vocabulary and grammar structures she would have obtained a great artefact whilst enhancing her students’ target language vocabulary and structural repertoire.

At the very outset of the project, on communicating to the students the project brief, teachers should be clear about the linguistic goals of the projects. Obviously, it is crucial to set realistic linguistic goals for the learners, as in PBL it is not rare to see students work at linguistic levels which are way beyond their current level of language competence.

  1. Plan for the integration of ALL 4 macro-skills and embed grammar and communicative functions

This point refers to another potential issue with PBL, the fact, that is, that often the medium one chooses for the project kind of drives the way the students process the language. Hence, if one decides to produce a movie to answer the big question that the project is meant to address, the students involved may simply focusing on writing a script and reading it aloud. However, as language teachers, we have the ethical imperative to forge a balanced linguist who is versed fairly equally in all four skills. Hence, teachers need to plan carefully for the integration of all four skills in the project.

As I have already pointed out in a past blog on digital learning, speaking and listening are indeed the two areas of linguistic competence that are usually neglected the most. Curriculum designers and teachers working in the PBL paradigm need to ensure that these two important macro-skills are not neglected in the process.

  1. Minimize ‘digital manipulation’ in the actual lesson

As I have already discussed extensively in previous posts (e.g. “Six useless things foreign language teachers do”), excessive ‘digital manipulation’ which occurs concurrently with language processing is likely to cause divided attention and will, consequently, impede learning. This is particularly the case with pre-intermediate to lower-intermediate learners. This phenomenon is due to the limited channel capacity of learners’ Working Memory who cannot attend simultaneously to various cognitively challenging tasks. Most ‘digital manipulation’ (e.g. App smashing) ought to be done by students outside lesson time, unless the presence of the teacher is absolutely necessary and the disruption to learning is deemed to be ‘minimal’.

  1. Emphasize language acquisition as the main goal of the project

The concern for the attainment of a well-manufactured finished product that PBL often – but not always – entails does on many occasions hijack the focus of the lesson away from the ultimate goal of MFL teaching, which is target language acquisition. This needs to be in the teachers and students’ focal awareness throughout the process, and tactics must be implemented to verify that each lesson is actually enhancing learner target language proficiency; hence, formative assessment must permeate the whole process and summative assessment needs to be implemented, too, at key moments. After all, PBL should be about the language and skills that are learnt in the process of carrying out the project, not about the final product.

  1. Make sure everyone in the group contributes

When PBL is carried out in groups some students often complain, at the end of the project, that not everyone contributed equally. It may be useful to require the students to keep a journal in which they must, at the end of each lesson, log in, in as much detail as possible, the extent of their contributions to the team’s effort. In my experience this is an effective way to scaffold fair and effective collaboration, especially if the journal is taken into account in the final evaluation.

  1. Control for unethical behavior

In this day and age, a lot of PBL will be conducted using the Internet. This increases the risk of plagiarism and online-translator use. At the outset of my past PBL experiences, I asked my students to sign an ‘agreement’ in which they pledged never to use online translator nor plagiarize any source during the project. Believe it or not, this symbolic act does usually pay off.

  1. Ensure the language assessment is valid and fair

For a test to be valid and fair, it must test students on what they have been taught. Hence, if we are teaching students through PBL, we need to be careful about how we test them at the end of the process. To simply assess them through the finished product is not valid, as the finished product does not tell us anything about how the process has enhanced their target language proficiency and about what they have retained. Nor can we test our students using traditional tests (e.g. the ones found in textbooks’ assessment packs), unless we have taught them through tasks similar to the ones used to test them. I have had colleagues in the past, who would do a project with their students on the same topic that the other instructors were teaching through different approaches whilst knowing that the end-of-unit test they would be sitting would have been based on the assessments found in the textbook-in-use (e.g. Expo, Tricolore, etc.). Not surprisingly, their students did not perform well.

  1. Provide clear guidelines as to how the project is going to be assessed

I put this last, as I believe this is obvious. However, a colleague of mine advised me to add this in as, in her opinion, this does not always happen. At the very outset of the project the students ought to be told how they are going to be assessed and should be talked through any rubrics or other evaluative procedures used to feedback on the final product. In my experience, the students should be reminded several times along the way of the criteria that will be used in the assessment in order to scaffold good quality and collaboration.

Six ‘useless’ things foreign language teachers do

images (1)

  1. Recasts

Recasts are the most frequent form of feedback that teachers give students in the course of oral interactions. They consists of utterances by the teacher that repeat the student’s erroneous utterance but ‘fix’ the mistake(s) without changing the meaning in any way. Example:

Student: hier j’ai allé au cinéma

Teacher: je suis allé au cinéma

Recasts, according to research (e.g. Doughty, 1994) are extensively used in the classroom representing up to 60 or even 70 % of all teacher feedback on oral performance. An interesting finding by Doughty is that recasts tend to concern themselves with minor errors rather than big problems.

As several studies have clearly shown, recasts do not really ‘work’ as they are not noticed most of the time. Havranek (1999) investigated to what extent learners recall corrective feedback from the teacher or their own or their peers’ mistakes. She found that less than one third of the learners who were corrected remembered having been corrected; peers did not pay attention to the correction of others and, most importantly, whether the corrections were recalled or not made little difference to whether the errors were or were not committed later.

The main reason why recasts do not work is that when the learners’ Working Memory is interrupted in the middle of speech production by the correction, it will not rehearse that correction for the time necessary to commit it to long-term memory -because it will be concentrating on resuming the interrupted conversation flow. Hence, the content of the correction will often be lost – which explains why Havranek’ subjects did not recall more than two thirds of the correction.

In view of the little surrender value of recasts in terms of acquisition, interrupting the students to correct them whilst they are talking may do more harm than good. Not only it may have a negative cognitive impact by disrupting their prospective memory; but it may also affect their self-esteem, especially if the correction relates to minor errors – as Doughty’s study found.

The above are valid reasons not to engage in recasts. It may be more productive and less threatening for the learners if teachers made a mental note of the mistakes noticed and treat them later on in contexts in which the learner’s attentional resources can be more productively channelled.

  1. Direct and Indirect error correction of written errors

Direct correction, whereby the teacher corrects an erroneous grammatical form and provides the correct version of that structure with an explanation on margin is pretty much a waste of valuable teacher time. Why? Tons of research (e.g. Cohen and Cavalcanti, 1990; Truscott,1996; Conti, 2001 and 2004) have demonstrated that most students do not process the correction in a way that is conducive to learning; most of them simply look at the mark, quickly read the comment and put the essay away, never to look at it again. Unless, as I argued in my post “Why teacher should not bother correcting errors in their students’ writing”, teachers engage students in more productive ways of processing teacher feedback, Direct correction is not going to enhance L2 acquisition. Error correction can be valuable when it places the errors into the students’ focal awareness, engages them in deep processing of teacher corrections, generates their intentionality to eradicate error and keeps it up for a sufficiently long period of time for any remedial learning to occur.

Indirect correction, on the other hand, is not likely to contribute much to acquisition as the learner will not be able to correct what s/he does not know (e.g. I cannot self-correct an omission of the subjunctive if I have not learnt it) and if s/he is indeed able to correct, s/he will not really learn much from it. To learn more about my views on this issue read my blog “Why asking students to self-correct their errors is a waste of time”.

  1. One-off learning-to-learn sessions

Not long ago I came across a beautiful Power Point on a teaching-resources website which purported to train students in effective approaches to the memorisation of vocabulary. It contained numerous slides packed with interesting suggestions on how to best commit vocabulary to memory and lasted long enough to cover a whole lesson. In the past, I myself produced similar Power Points and delivered one-off sessions on learning strategies which the students usually found quite interesting and engaging. But did they actually learn from them?

The problem is that, unless there are several follow-up sessions and some form of scaffolding reminding the students to use the strategies that the Power Point presented, thirty years of research (see Macaro, 2007) clearly show that this approach does little more than raising learner awareness of the existence of these strategies, but will not result in learner uptake, i.e. very few if any learner will incorporate these strategies in their active repertoire of learning strategies.

For any learner training to be successful it must involve learners in extensive practice of the target strategies.

  1. Identifying students’ learning styles and planning lessons accordingly

Research has clearly shown that learning styles and multiple intelligences are invalid constructs totally unsupported by theory and research. Moreover, there is not a single shred of evidence to show that teaching students based on their alleged learning style actually enhances their learning. Teachers should not waste valuable teaching time administering questionnaires or other ‘tests’ in an attempt to identify students’ learning style or ‘dominant intelligence(s)’. Most importantly, they should not bother planning lessons or remedial learning programs based on the findings obtained.

In view of the invalidity of these constructs, labelling students as visual, kinesthetic or other may lead them, especially the younger ones, to form a self-fulfilling prophecy that may ultimately be detrimental to their learning.

  1. Asking pre-intermediate/lower intermediate learners to peer assess oral performance

Although it has some (modest) surrender value in terms of metacognitive enhancement, the practice of involving fairly inexperienced learners in peer assessment is not justified by the learning gains it produces, especially in terms of language acquisition. Firstly, as these learners do not usually possess enough declarative knowledge of the language to be able to assess and feedback on language use in a way that can significantly benefit the recipient of the feedback; secondly, and more importantly, they do not possess sufficient levels of procedural knowledge to be able to apply any declarative knowledge they have whilst processing what they hear their classmates say – which means they cannot effectively evaluate their oral output.

In fact, even with more proficient learners peer assessment practice may not always be beneficial. In a little experiment I made last year, I got 16 students that I had practised peer assessment with almost on a daily basis to assess their classmates after a typical IGCSE conversation, using the CIE evaluation rubrics. When I compared their assessment scores to mine the discrepancies were huge, most of them having been on average 25 % more generous than me in allocating marks.

  1. Asking students to create digital artefacts in class

As I wrote in my blogs ‘Five central psychological challenges of foreign language learning’ and ‘Of SAMR and Samritans”, creating a digital artefact in class is not likely to be conducive to language acquisition enhancement at pre-intermediate to intermediate levels of L2 proficiency. The main reason is that human cognitive resources being finite, the working memory of an intermediate/lower intermediate MFL learner will not usually be able to process language effectively and efficiently whilst concurrently focusing on the operations he/she will be performing, e.g. cutting, pasting, ‘googling’ pictures, videoing, recording, ‘smashing’ Apps, etc.

Hence, forgetting by divided attention often occurs with not much learning at all taking place. I will never forget a group of Year 6 students telling me, after being involved for 5 weeks in making an iMovie about the topic ‘Ma maison’ in class, that all they remembered was the French for the rooms in the house.

Students, when involved in such activities should not do any ‘digital manipulation’ in lessons, unless we believe that this is very likely to enhance their target language proficiency. Classroom time should be devoted to learning the target language.

You can find out more about my approach to language teaching and learning in the book I co-authored with Steve Smith: “The Language Teacher Toolkit” available for purchase at http://www.amazon.com / http://www.amazon.co.uk

‘Noticing’ the L1-L2 gap in the foreign language classroom – how it can enhance learning


Schmidt’s (1990) so called ‘Noticing hypothesis’ proposes that noticing the ‘gap’ between the first and the second language is a powerful catalyst of language acquisition (see picture above). For example, if an English learner of French notices that ‘I have gone’ in French is ‘Je suis allé’, he will realize that the auxiliary ‘Etre’ (=to be) is used in the target language instead of ‘Avoir’ (= to have) to form the Perfect tense of ‘Aller’ (= to go); this will trigger the process of acquisition of the perfect tense of ‘Aller. The cognitive comparison between the L1 and L2 versions of grammatical, lexical or phonological structures across the key areas of linguistic competence that noticing involves would, in other words, spark off the target language acquisition process.

Schmidt’s theory is plausible, as foreign language learners always use the first language, or any other languages they are expert in, as a starting point for inferences as to how the target language system works. I have often experienced this phenomenon in my own language learning experiences.

I would add to ‘noticing the gap’ the importance of ‘noticing the similarities’ between the first and the target language, too, in L2 acquisition. This is as crucial as noticing the gap, in that the foreign language learner needs to understand the extent to which the L1 and the L2 systems overlap and L1 positive transfer may be used advantageously. In other words, noticing both the gap between the languages and their overlap help synergistically the MFL learner in gaining control over the way the L1 system works.

Schmidt (1990) views noticing as the most crucial factor in promoting language acquisition. I do not entirely agree, as I can think of several other routes to acquisition which do not involve noticing. However, I do believe that noticing is an important teaching and learning strategy which can promote the acquisition of declarative knowledge of the foreign language’s rule system.

With this in mind, one would expect noticing to be used quite frequently in the typical MFL classroom, either (a) as a starting point for the teaching of grammar, syntax, pronunciation, lexis (especially idioms and collocations) and sociolinguistics, or (b) to explicitly model it as a learning strategy that students may use to their advantage. However, this does not, in my experience, happen that often. Yes, teachers do allude to the differences and similarities between the L1 and the L2 in passing during lessons; however, the full learning potential of this powerful learning catalyst is often not tapped into sufficiently and not across all the key areas of MFL learning.

Let us consider pronunciation, for instance. As I have already pointed out in a previous post, raising awareness of the differences between the L1 and L2 phonological systems should be encouraged from the early stages of learning. Yet, this practice is often overlooked, according to the feedback I have had from many readers, especially in UK schools. To expect all of our students to spot, focus and consciously act on the differences between the two languages is naïve, to say the least. In fact, contrastive phonology is one of the most neglected MFL teaching strategies in this day and age, yet, is one of the most effective ways of improving learner pronunciation, especially when ‘poor’ pronunciation has been fossilized. Playing a recording of the same L2 sentence as pronounced with a native/near-native accent and subsequently with a typical (fairly strong) foreign accent and ask the students to spot the differences is an easy activity to set up, the students enjoy it and it is an excellent pre-speaking task (when one wants to focus the learners on pronunciation and not simply fluency).

Another context where the neglect of the learning potential of noticing and L2-L1 cognitive comparison is evident is reading. As I have argued in a previous post, texts are usually ‘read’ in the context of comprehension tasks that feel like tests; once the reading comprehension is over and done with, the text is ‘ditched’ and nothing else is learnt from it. Yet, providing the students with the translation of that text after completing the comprehension tasks, substantially enhances its learning potential, as it allows the students with a powerful tool to notice many differences between the languages at different levels: grammar, syntax, use of idioms, use of genre-specific conventions and even punctuation. I often give such parallel texts to my students and challenge them to find 10 differences across all the afore-mentioned categories; they enjoy it and it sparks off a lot of interesting metalinguistic conversations (when the students work in groups) whilst enhancing their inductive/inferencing skills.

Correction, is another ideal context for promoting ‘noticing’; but how many times do we use it to raise awareness of the ‘gap’ between the L1 and L2 word order, use of relatives and determiners, moods and tenses, etc.? Yet, in one of my studies, my informants reported learning more from error correction when the teacher had made explicit reference to the differences between L1 and L2 usage of a structure or idiom and/or had asked them to reflect on and explain such differences.

Another example of the benefits of promoting noticing is grammar instruction. However, textbooks or teachers in my experience rarely present L2 grammar examples of the target structures with the L1 translation next to it making explicit reference to the L1-L2 differences in their usage. Yet it is easy to do and with the effective use of visual aids such as colour coding and other high-tech gimmicks the translation is very likely to enhance students’ understanding of the usage of the target structure. Moreover, noticing can be used to great effect in the context of inductive grammar tasks, whereby the students are provided with several sentences (with the translation aside) containing examples of the target structure and are asked to work out the rule(s) governing that structure. In the context of this type of activities, ‘noticing the gap’ acts as a cognitive facilitator and as a useful springboard for metalinguistic inferences.

Spelling, too, can benefit from cognitive comparison and noticing. Let us consider cognates, which, especially when they are very similar in the two languages, can confuse the learners in written recall (cross-association effect) – I have had a few painful reminders of how frequent this phenomenon is today, whilst correcting an essay written by a very good student. Focusing on the differences between cognates at the level of spelling and ‘drumming them in’ through a number of activities such as gapped words, odd one outs – where three versions of the L2 cognate are given and only one is correct – , hangman, jolly old fashioned dictation, etc. can address this issue effectively.

Essay-writing is another important area where noticing the gap between the two languages can be of great advantage. MFL teachers – unlike EFL teachers – rarely, in my experience, use essays written by native writers to model the specific genre conventions and other discourse features typical of L1 essay writing. Students are expected to write essays the way they were taught to write in their L1. But a French native writer would not develop an essay like an English native writer would. Nor would the sentences written by the former be as short, concise and simple as the latter. The formulae used to open and close essays would be different, too. Etc.

Translation tasks from the L2 to the L1 are another great source of noticing, especially if the students are asked, as a follow-up task, to identify a set number of cases – like I suggested for parallel texts earlier on – where idioms, grammar, syntax or lexical collocations differ across the two languages. This is also not done often enough, at least not in UK MFL classrooms.

But if noticing can be beneficial for foreign language acquisition, why is it not explicit encouraged that often, especially through the kind of activities I envisaged above? As the insightful Steve Smith of www.frenchteacher.net wrote to me once, when one believes that lessons should be entirely or almost entirely in the target language, the use of the first language becomes almost taboo. Moreover, noticing is a phenomenon often taken for granted; teachers presume that there is no need to make the comparison between the two languages explicit, as students will do it anyway; the issue is to move this process, when it does occur, from the learners’ subsidiary to their focal awareness. Unfortunately, a lot of students are not as foreign-language savvy as we may think and noticing may require levels of metacognition that not all the students possess.

In conclusion, noticing should be encouraged and modelled more often and systematically than it is currently done, across all the key areas of MFL teaching and learning. One great benefit for learning that I have not mentioned thus far is the enhanced awareness that a language is not a word-for-word translation of another. This enhanced awareness may wean our learners off online translators which create linguistic monstrosities that upset a lot of us almost on a daily basis. An activity I do at least once a term with my students is to provide a number of (carefully constructed) English sentences with their French translation; ask them to notice and highlight the differences and finally turn those sentences into French using Google translator to make them aware of the errors stemming from L1-to-L2 word-for-word translation. An effective awareness-enhancer and/or reminder that there is more to language learning that substituting every English word in a sentence with its Spanish, French or German equivalent.

Five central psychological challenges facing effective mobile learning

boy with ipad

A while ago, a very good friend and former colleague of mine, Fiona Seymour, popped around my classroom to have a chat with a group of year 8 students about how they felt the use of the iPad had impacted their learning since its adoption in our school. Some interesting facts emerged from the discussion which referred to some cognitive and metacognitive challenges that learners face when using the iPad or any other mobile device for learning.

Interestingly, in a very informative article that Fiona shared with me a few months later, two Scottish researchers from The University of West Scotland, Dr. Melody Terras and Dr Judith Ramsay address the very same challenges that my students mentioned, recommending that education providers take them into account when adopting mobile learning as the main or one of the main modes of delivery of the curriculum.

As a fan and assiduous user of mobile technology in the classroom I found this article a true ‘eye-opener’ as it not only confirmed some of the concerns I had about mobile learning but also triggered an ongoing process of reflection on their possible implications for teaching and on what I could do in my daily practice to address those concerns.

I believe than at an exciting time where ICT integration in the classroom is a reality in most developed countries every educator must be aware of these psychological challenges and endeavour to address them consistently and systematically in their teaching.

The five psychological challenges

Terras and Ramsay (2012) (at http://onlinelibrary.wiley.com/doi/10.1111/j.1467-8535.2012.01362.x/abstract ) identify the following crucial challenges of mobile learning:

  1. The context-dependent nature of memory

As I have already discussed in a previous blog (“Words in the mind – how vocabulary is stored and organized in the brain”), memory is context-dependent. In other words, the context in which we create a memory will enhance our chances to recall that memory later. Thus, for instance, if I am learning a set of new vocabulary in my classroom, whilst sitting next to my friend Joe and facing my teacher, who is wearing a bright red flowery shirt, the classroom, my friend Joe, my teacher and her shirt will enhance my chances to recall that vocabulary in the future. Memory is also state-dependent. In other words, individuals are more likely to effectively recall a memory when he/she in the same emotional, motivational or physiological state in which he/she was when he/she encoded that memory.

As Terras and Ramsay point out, research shows that the retrieval of recently learnt material is highly affected by the influence of context. These findings have huge implications for mobile learning, as mobility, the fact that the learner may be using the iPad or phone to study in different environments, may disrupt the support of the context as a cue for retrieval of the target information. Thus, for instance, a student who used the iPad to learn new vocabulary for a test in his room, may then be revising it in the car or school bus on the way to school,then, once in school, may be going through it again in the library or in the canteen; each change of environment being different, the context-related retrieval cues will be missing, with a possible negative impact on recall.

Hence, the main learning-enhancement advantage of mobile devices, i.e. the fact that one can carry them with oneself wherever one likes, has the potential to disrupt learning. What can teachers do to address this issue? Terras and Ramsay do not make any pedagogical suggestions. My take on this is that, like one should do with any learning tool, teachers have the ethical imperative to forge healthy learning habits; this entails (a) raising learner awareness of this issue and of how it can affect learning and (b) modelling effective memory strategies which may effectively compensate for the lack of context-based cues; for example, students may be taught that, on learning new vocabulary whilst reading a French text on the iPad, they can creatively associate them with images and L1 words through the so-called ‘Keyword technique’, index cards and other menmonics or by using Apps like ‘Poplet’ or ‘Padlet’ to store and organize it semantically during or after reading the text. The association created through these two approaches would function as powerful retrieval cues at recall.

  1. Human resources are finite

As I have often reiterated in my posts, one of the greatest ‘enemies’ of learning, if not the greatest, is divided attention caused by cognitive overload on Working Memory (e.g. processing inefficiency) or interference (e.g. being distracted by an another stimulus whilst trying to learn). As many of my students have pointed out over the years, mobile devices can generate a lot of distraction, mainly coming from notifications from Facebook, Instagram and e-mail. Terras and Ramsay report a study by Cicso investigating the ability of social media to distract students. The study revealed that UK students are ‘most distracted by social media’. Another sort of distraction that Terras and Ramsay do not consider, but that I notice day in day out in my school and surrounding areas, is that mobile devices allow for students to study whilst socializing (e.g. they sit around in groups each working on their iPad). This was less likely to happen with PCs and laptops. In sum, as Terras and Ramsay (2012: 824) advocate,

Mobile learners may need to be more skilled at inhibiting responses to extraneous stimuli. In particular, they may need to develop superior attentional control in order to be effective learners in environment that are not primarily designed for learning. Noisy and changing environments and the potential distractions posed by social media may place significant additional demands upon the learner’s auditory and visual attention.

As they also point out, one important harm that interruption causes to learning occurs at the level of prospective memory. Prospective memory supports the intended execution of future tasks and may be time-based (e.g. remembering to check an essay once more first thing in the morning the next day before handing it) or event-based (e.g. remembering to arrange ideas in a logical sequence when planning an essay). Disruption seems to have a particularly harmful effect on prospective rather than retrospective memory, especially on the stages of prospective memory involving execution and evaluation.

Divided attention can stem from the environment but also from the mobile medium itself. As Terras and Ramsay point out, research evidences that face-to-face communication always outshines mediated communication as the latter is usually less rich and more ambiguous. Mobile devices, websites and apps pose more cognitive challenges, especially for less able and flexible learners who may experience processing overload whilst accessing and/or manipulating it.

Another important harmful effect was highlighted by myself in a previous post on the SAMR model (“Of SAMR and Samritans…”) and refers to another powerful source of divided attention: the cognitive load posed on Working Memory by the ‘mechanics’ involved in the creation of a digital artifact. Often teachers involve students in tasks which require them to operate simultaneously on two levels: on the one hand they are required to generate, elaborate and organize ideas related to a given task’s brief, on the other they are required to convey the resulting intellectual product of that process through digital media (e.g. smashing apps). If teachers are not careful, the demands posed by such modus operandi can easily cause cognitive overload and impede learning.

Sources of cognitive overload must be anticipated and addressed in the planning of any activity in which digital media are integrated with MFL learning.

  1. Distributed cognition and situated learning

By this, Terras and Ramsay refer to the new phenomenon created by the Web, whereby learning contexts which in the past where very distant from one another are becoming increasingly interconnected via mobile social networking. Thus, mobile learners construct their comprehension of the world and knowledge through cognitive interaction with a much greater and more culturally diverse range of contextual sources of information than non-mobile learners. This has obviously the potential to greatly enhance learning. However, the challenge resides in the fact that not all individuals, nor all external input, will be of value for the learning process. Hence, mobile learners need to be ‘taught’ how to discern who and what on the web is relevant to their learning and reliable as a source. As Terras and Ramsay put it:

Learners will have to cope with an extra layer of complexity in their learning ecology: mobile social learning increases the density of the distributed cognitive network. So, although learners may benefit from this increase in the distributed and situated nature of their cognitive ecology, the challenge is to use their digital literacy skills in addition to more generic cognitive skills in order to screen out redundant or irrelevant input to their learning

Education providers, in my opinion, need to invest much more than they usually do in structured learner training programs aimed at raising learner awareness of this challenge, whilst modelling effective approaches to a safe, discerning and cognitive-/time-efficient use of web sources. Such programs should be implemented, in my view, concurrently and as extensively and intensively as digital literacy programs are – which, as research suggest, rarely happens.

In the MFL classroom this is an important issue across several dimensions of learning. Firstly, mobile MFL learners are more likely to be tempted to use online translators and must be warned about the dangers of their use. Secondly, they need to become more discerning as to the cultural or political bias of the target language sources they interact with. Thirdly, they must be able to grasp the differences in terms of register, between different types of text genres (e.g. how writing a facebook feed differs from writing a blog or journal article). Fourthly, plagiarism is highly encouraged by the mobile social networking culture where information is recycled at high speed with little regard for intellectual property.

  1. Metacognition is essential for mobile learning

This is undoubtedly the most important challenge. Mobile learners must possess the ‘psychological infrastructure, as Terras and Ramsay put it, to support mobile learning. In other words, they need to develop metacognitive skills to be able to cope with all the above mentioned challenges. They need to develop strategies in order to best prevent the already discussed sources of cognitive load and external distractions from interfering with their learning and to generally effectively self-manage the learning process; this is particularly important if we aim to develop truly competent autonomous learners. The development of mobile learners’ web-related metacognitive competence should start concurrently with the very early stages of mobile learning.

  1. Individual differences matter

Although Terras and Ramsay consider this as a different point to the previous one, I believe the two are  very closely related. What the authors mean here is that students need to understand how technology best suits their personality, age, gender, learning preferences, personal set of skills, aptitudes and attitudes. Teachers often take it for granted that mobile learners, simply because (in their daily life) spend hours on mobile devices will know how to use them for learning. Being able to use mobile apps, social and learning platforms in a way that best suits one’s own personality attributes, skills and academic goals is not easy – it is a complex skill to acquire. How many teachers, I wonder, have the know-how to effectively impart on their mobile learners training in this kind of competence? This implies that professional development in this area ought to be focused on by education providers so as to equip teachers with the necessary knowledge and skills to address this crucial aspect of learners’ metacognition.

I would add a sixth psychological challenge to the five that Terras and Ramsay have identified: Depth of processing. Young mobile learners are exposed in the social media (e.g  Facebook and Twitter) to an overload of information, much of which is ‘fast’, ‘sloganised’ (forgive my ‘neologism’, here) and tend to be ‘sensationalistic’ and ‘eye-catching’ in nature. This ‘high-impact’ / ‘high speed’ culture has created a mindset amongst youngsters which encourages a superficial approach to information and cognition. This mindset, in my view, engenders shallow processing and the skin-deep acquisition of facts and notions unsupported by substance and/or reliable and referenced evidence. The challenge is for teachers to engage students, through mobile learning, in deeper processing of facts, notions and ideas. It is no easy challenge in a fast-paced society like ours, where the digital world creates on a daily basis such a wide and divergent pool of information and entertainment opportunities. We need to always bear in mind, as educators, that it is depth of processing that, after all, creates effective learning.

In conclusion, mobile learning has an enormous potential for the enhancement of MFL learning and of learning in general. This potential, however, has to be effectively harnessed and channelled. Education providers have to be aware of the 5 challenges pointed out by Terras and Ramsay as they are central to the construction of our learners’ cognition and may affect learning. In a highly interconnected world where social networking dictates a fast-paced, scarcely regulated and rich flood of information, the 21st century mobile learners need to be trained effectively to become competent autonomous learners versed not just in digital literacy but, more importantly, in generic life-long skills which enable them to analyze, synthesise and evaluate judiciously, productively and safely the masses of information available online.

Narrow reading and narrow listening – enhancing receptive skills through focused and purposeful recycling


As mentioned in previous posts, listening in my opinion is not taught effectively in many MFL classrooms as students are often engaged in listening comprehension tasks that ‘feel’ more like tests and involve almost exclusively top-down processing skills. Another reason is that instructors and textbooks often do not exploit the full potential of the target texts. Students listen to a recording two or three times, answer a few comprehension questions and then the text is ‘ditched’ and the teacher moves on to a new text  which may deal with a similar topic area but rarely recycles the same material. As far as reading is concerned, things are a bit better and occasionally textbooks do provide two or three activities on the same text. But even so, the full potential of a text is often wasted.

In what follows I propose an approach to listening and reading that I have used very often over the years, which, although quite time-consuming, can yield great results if executed properly and implemented regularly.

Narrow listening

When I was in my teens, Internet did not exist and finding the lyrics of a song, in Italy – where I lived in those days – was not easy. Hence, a lot of my school mates, aware of my anglo-phone background (my mother having grown up in England) would ask me to write up the lyrics of the British hits that were ‘invading’ and taking over Italy at that time. To be able to do that, although I was already near-native at that stage, I had to listen to each song over and over again, sometimes going over the same line a dozen times to decode problematic words.

Useless to say, that process was as useful as it was tedious; I learnt loads, both in terms of bottom-up processing skills and in terms of vocabulary. But can we ask students to do the same? Apart from a few highly self-motivated individuals, the vast majority of our learners would not. Yet, as I argued in a previous post, if students simply listen to a text once or twice only, the learning benefits will be relatively low; students often need to process the same text several times in order to fully understand and learn new vocabulary from it.

One obvious solution is to exploit each text several times by engaging the learners in four or five tasks based around it.  However, this can be boring and repetitive. So, what can one do, to ensure that the students listen to the same words over and over again without listening and re-listening to the same text?

I found a possible solution a few years back, whilst reading the work of Stephen Krashen and his ‘narrow listening’ technique. ‘Narrow listening’ involves asking several L2 proficient/native speakers to talk about a specific topic whilst recording them in the process; the questions should be quite ‘narrow’ in their focus so as to elicit similar content and, consequently, language (vocabulary and grammar). At the end of the process one would listen to all the recordings obtained thereby being exposed several times to fairly similar language.

When I first got acquainted with this technique, I liked its aims: firstly, to facilitate learner understanding of the target input, by creating the same ‘narrow’ context for each interview and by recycling the same vocabulary over and over again; secondly, to consolidate vocabulary through that recycling. However, Krashen’s approach, being based on a spontaneous response on the part of the interviewees, does not guarantee any control on the teacher’s part over the vocabulary contained in their input. This means that the effect that song-transcribing had on me could not be guaranteed 100 % all of the time through Krashen’s technique.

Hence I decided to adapt Krashen’s model and increase my degree of control over the input. In my model, the interviewees are not just improvising; each is given a script that will not take longer than 30 to 40 seconds to read for beginner to pre-intermediate learners and about 1 minute for GCSE level students. The wording of each script is quite similar – but not identical – and the input is ‘comprehensible’ (i.e. mostly familiar language or cognates, with a few unfamiliar words). The ‘secret’ is to make sure that the same pool of words is recycled constantly whilst the texts sound different and have slightly different messages (e.g. some negative; some positive; some ‘neutral’) – not an easy thing to do. Even though the process can be quite laborious, the learning benefits of this practice in terms of vocabulary acquisition, consolidation and self-efficacy are remarkable.

In international schools like the one I work at where there are lots of young native speakers of the target language(s) (and their parents) this approach is not too difficult to implement. But one can go about it a different way, by searching the web for short videos/recordings which are very similar in content. I did that, for instance, in the context of daily routine; I found a serious of youtube videos shot by French teenagers, which actually contained very similar language and used them for narrow listening tasks. In the absence of L2 native speakers, other L2 experts (e.g. colleagues) may be used.

But what tasks should the students be involved in whilst narrow-listening? Unlike Krashen, who believes they should not be doing anything but listening, I do believe the students should be demonstrating understanding one way or another. The type of activity we decide to engage them in will depend on what we are trying to focus them on: is it inferring the meaning of unfamiliar vocabulary? Is it consolidating old vocabulary? Is it listening for gist? Is it identifying specific details? Or is it all of the above? The answers to these questions will determine the in-listening tasks we will set. I usually do one or more of the following:

  1. Given the word in English spot the French/Spanish equivalent on the recording;
  2. Identify the meaning of the following French words in the recording;
  3. Jot down two/three/four etc. main points each speaker makes;
  4. Spot the speakers who express similar/different opinions or do/have done similar/different things;
  5. Comprehension questions. As you will know if you read my previous posts, these are not my favourite kind of tasks.

Narrow listening activities should be preceded by warm-up vocabulary and schemata activation activities and followed by consolidation activities further recycling the target lexis in order to maximize retention.

Narrow reading

The same principle underlying narrow listening can be applied to reading with the same beneficial impact on L2 acquisition. In fact, I often use the transcripts employed in narrow listening for my narrow reading sessions. I strongly recommend using fewer texts for narrow reading than one would use in narrow listening. Narrow reading is easier to implement as there’s no need for native speakers and similar texts are easy to find. For instance, recently I was doing some work on the environment and finding five similar texts on things to do or not to do to protect the environment was easy – and each text was authentic L2 material.

In conclusion, narrow listening is a technique that I recommend to colleagues and that I wish textbooks and other published instructional materials adopted more often than they currently do. It has the great advantage of exposing the students to similar comprehensible input, which allows for easier access to unfamiliar language, due to the contextual and linguistic clues they get from listening to several similar texts; moreover, similar vocabulary is recycled over and over again which fosters consolidation and retention; finally, students get to listen to accessible L2 language for relatively long time as uttered by different people. However, it does require some extra-work and the availability of several L2 experts willing to co-operate.

In order to save time and effort, Krashen’s approach may be easier to implement and can work, too, if one chooses the right questions. I do it sometimes and ask students to go around school with their iPads – not in lesson time, obviously –  to interview as many L2 native speaker schoolmates as possible about a specific topic (e.g. food in the canteen). When the focus of the question is very narrow (e.g. talk to be about a typical day in school), the likelihood of the vocabulary overlapping across interviewees is quite high.