Why do our learners often get prepositions, articles and verb endings wrong?


This article was prompted by a question (the one in the title) a colleague asked me recently at finding lots of mistakes in their students’ essays relating not solely to prepositions, but also to definite/indefinite articles, copulas (e.g. is and are) and other function words. The answer to that question is relatively simple, but in order for the reader to fully grasp its implications for classroom instruction, one has to first get acquainted with the concept of Working Memory and Executive Control.

Working Memory

To put it as simply as possible – since one of my colleagues keeps complaining about the complex jargon in my blogposts – Working Memory is the space activated in our brain when we process information, what in the old days was called Short-Term memory. Working Memory is a ‘buffer’ between the outside world and Long-Term memory which ‘holds’ any information we are trying to decode and retrieves the information we need from Long-term memory which we need to carry out the task-in-hand (e.g. writing a sentence or understanding a text). So, for example, when we are writing a sentence in a foreign language, Working Memory is the ‘place’ along our neural network in which we actually construct that sentence (i.e. where we choose the words we need from our mental lexicon, arrange them together in a grammatically correct sequence, make sure the spelling is correct and edit the final product).

Working memory has very limited channel capacity, in other words can only store a limited number of images, words and numbers at any one time and unless we keep rehearsing it, the information will be lost easily after a few seconds only (a word stays in our brain only 2-3 seconds unless we make a conscious effort to retain in through rehearsal). That is why, in order to keep a phone number in our head as we frantically try to key it in our phone we need to repeat it in our heads (or rehearse it) a few times. Miller’s (1965) magic number 7+/- 2 indicates the number of digits we can hold in our Working Memory at any one time – a very short number indeed.

The challenges posed by foreign language writing

Writing in a foreign language is much harder than a lot of us may think, especially under communicative pressure. Let us have a closer look at how a sentence is produced in writing. First of all, it is important to point out that the starting point, both in the first and the second language is a Proposition, in other words a representation in our brain (in Semantic Memory to be precise) of the concept or idea we are trying to convey. A Proposition, unlike what we may intuitively think, is not made up of words, thus, the brain has to translate into words, whether we are operating in the first language or second language.

According to Cognitive research (e.g. Cooper and Matsuhashi, 1983), the Translation process consists of four stages: Wording, Presenting, Storing and Transcribing. In the first stage, the brain transforms the Propositions into words (lexis). Although at this stage the pre-lexical decisions the writer made at earlier stages and the preceding discourse limit lexical choice, Wording the proposition is still a complex task: ‘the choice seems infinite, especially when we begin considering all the possibilities for modifying or qualifying the main verb and affected nouns’ (Cooper and Matsuhashi, 1983: 32).

Once s/he has selected the lexical items needed, the writer has to tackle the task of Presenting the proposition in standard written language. This involves making a series of decisions in the areas of genre and grammar. In the area of grammar, Agreement and Tense will be the main issues, especially in languages like French, or German where a lot of permutations are required.

The proposition, as planned so far, is then temporarily stored in Working Memory while Transcribing takes place. Propositions longer than just a few words will have to be rehearsed and re-rehearsed in Working Memory for parts of it not to be lost before the transcription is complete.

The limitations of Working Memory create serious disadvantages for unskilled writers. Until they gain some confidence and fluency with spelling, their Working Memory may have to be loaded up with letter sequences of single words or with only 2 or 3 words (Hotopf, 1980). This not only slows down the writing process, but it also means that all other planning must be suspended during the transcriptions of short letter or word sequences.

The physical act of Transcribing the fully formed proposition begins once the graphic image of the output (what the sentence physically looks like) has been stored in Working Memory.

In L1-writing the decisions taken at any of the four stages outlined above are taken automatically, thereby occupying little or no space at all in Working Memory. However, in L2-writing, especially in beginner to intermediate writer, every decision will take a lot of Working Memory space, making the process slow, cumbersome and difficult to monitor because the process happens mostly consciously.

Hence, the adaptive response of the brain, especially in beginner writers, is to prioritize the most important features of each proposition (the principle of ‘Saliency first’ being at play here), i.e. : the items that are most important in terms of conveying the intended meaning. The most semantically salient elements will include mainly: Nouns, Verbs and Adjectives. Function words, which carry considerably less meaning, will be relatively neglected by Working Memory’s attentional systems as, let’s face it, even if the writer gets them wrong, they won’t impede comprehension massively (example: whether I say, in French “Je vais au cinema’ or ‘je vais à la cinema’ I will be readily understood by a reader/listener).

This phenomenon is exacerbated by linguistic distance between the first language and the target foreign language. For instance gender (masculine and feminine) as well as verb endings are not likely to be perceived as salient by an English native speaker (as they do not exist in their language), which means that they are likely to be less monitored.

The less proficient the foreign language writer is and the less time he has to monitor his/her output, the more likely he/she will be to make mistakes with function words. Hence, errors are bound to be even more frequent in oral performance, where the self-monitoring capacity of Working Memory is drastically reduced compared to the written medium.

As the learner becomes more proficient, his/her ability to juggle the demands posed to his/her Working Memory by the processes outlined above will increase. This is due to the fact that with a lot of writing practice in the target foreign language a lot of sub-processes become automatized and require only peripheral attention, freeing up Working Memory space. This enhanced processing efficiency will also allow for more accuracy, too, in the production of less salient features unless Error Fossilization throws the spanner in the works.

The danger of fossilization

When errors go unmonitored a bit too often, they become automatized and it is very difficult to ‘unlearn’ or eradicate. Mukkatesh (1986) found that despite many remedial interventions such errors cannot be eliminate at all from L2 learners’ Interlanguage. This phenomenon, called by Selinker (1972) Fossilization, is obvious in a lot of foreign language speakers, especially when it comes to pronunciation; that is why, according to Selinker, only 5 % of foreign language speakers can be said to sound 100% native-like. Their second language will always contain some fossilized item. A very good friend of mine, for instance, speaks perfect English, with accurate pronunciation and grammar and an impressive lexical repertoire wider to that of an average native speaker; however, he cannot help voicing the ‘p’ in the word ‘psychology’ (influence of his first language: Italian), despite many corrections. Such is the power of Fossilization.

Function words and any other less salient L2 features (e.g. gender, plural and verb endings and minor pronunciation inaccuracies) are particularly amenable to fossilization as they are more likely to go unmonitored and uncorrected. Therefore, the danger is that when learners do not get enough negative (cognitive) feedback at the early stages of L2 acquisition, they are likely to fossilize mistakes with the above L2 structures and to keep making these mistakes all the way to A-Level and university – as I have often witnessed in my university lecturer days.

Communicative language teaching, especially in its strong version, by prioritizing fluency over accuracy, often leads to fossilization (and pidginization) especially when the students are asked to perform in unstructured oral practice, at a level of proficiency they are not developmentally ready for and under communicative pressure. (Skehan, 1994)

Implications for MFL teaching


The implications for the MFL classroom are manifold but hinge mainly on the teacher’s pedagogy and on the course end-goals. If we are teaching GCSE level students and we are happy for them to make a few minor mistakes as far as they can convey their intended meaning effectively, we should not worry too much about error and we can exercise a relatively high degree of tolerance. However, if we are dealing with individuals who want to make language their career and become one day interpreters, translators or teachers, then the attitude has to be less lax and mistakes with articles, prepositions, copulas and gender agreement WILL matter.

If we do want to address this issue radically, we need to keep students focused on the importance of accuracy from the very early stages of language acquisition whilst keeping the main focus of our teaching on the development of fluency. This is not easy, even for experienced teachers. Editing instruction – through games, quiz and other fun activities – should become part of almost every lesson (through snappy starters or plenaries, for example) to remind students of the importance of accuracy and to raise their awareness of which mistakes are more likely to occur at their current level of proficiency.

More importantly, the written tasks we involve are students in must pitched to the correct level, especially in terms of the cognitive challenges they pose to an inexperienced writer. If we do not, we are likely to engender more error than we and the students can effectively deal with in the remedial phase. Fluency, as I said above, has priority, it is true; however, fluent output that is rife with errors can be stigmatizing and irritating for the reader/interlocutor and we need to be aware of that in a global era in which, more than ever before, our learners are more likely to use the target language in the workplace.

Finally, Error correction – or rather Error remediation – can also play an important role if it engages the learners in a sustained long-term self-monitoring process initially moderated by the teacher which aims at focusing them on their most frequent mistakes.


How to lessen the negative interference of our learners’ mother tongue on their target language pronunciation


This article suggests teaching strategies based on sensory-motor research findings on bilingual speech aimed at reducing the impact of negative first language transfer on L2 target language speech. I will start by examining the process involved in language transfer and its consequence on the acquisition of L2 pronunciation in light of the latest recent research evidence. I shall then proceed to discuss the implications for the foreign language classroom.

Language Transfer

As we all know, our first language (or any other language we know for that matter) can cause interference in the process of acquiring a new one. We refer to this phenomenon as language transfer. Language transfer can be positive (i.e. facilitating learning or performance) or negative (i.e. impeding learning or performance) depending on the similarity/distance between the pre-existing language and the new language one is learning.

For instance, an English native speaker will experience negative transfer when it comes to pronouncing ‘P’ in Spanish ( negative transfer), because in this language the ‘p’ sound is a non-aspirated labial sound with a short onset time, whereas in English it is an aspirated sound with a relatively long onset time. On the other hand, a native speaker of Italian will have no problem with that sound as in his/her language it is pronounced in exactly the same way as it is in Spanish (positive transfer)

The main cognitive cause of Language Transfer is that when we learn a new language our brain uses our default language(s) – more than often our first language – as the starting point for the hypotheses we formulate to make sense of that language and/or as a communicative strategy to fill in any communicative gaps. In the specifics of foreign language pronunciation L2-learners transfer refers to the L2-learners’ application of their L1-phonological categories to decode and represent the foreign language sound system. This phenomenon is exacerbated by the fact that their motor commands (their control over larynx, phrarynx and articulators) have been conditioned by years and years of first language pronunciation. Hence, especially at the early stages, the ‘phonological distance’ (differences in pronunciation) between two languages will play a very important role in determining the accuracy of L2- learner pronunciation.

Negative transfer is more likely to cause error at pronunciation level, when speech occurs in contexts that are difficult to monitor or which require a greater mastery of motor skills. So, for instance, a beginner foreign language learner talking spontaneously in the context of uncontrolled communicative practice will have less time to monitor pronunciation because his/her Working Memory is focused on higher metacomponents such as meaning and grammar; in this kind of context, the target language sounds that s/he finds problematic will be seriously affected by the lack of monitoring. On the other hand, the pronunciation of those problematic sounds is usually much more accurate when they are uttered in isolation, as discrete items – just like a toddler’s blabbing – due to the absence of articulatory interference from the preceding and following sounds in the word/surrounding words and ease of monitoring.

Another way in which L1-transfer affects pronunciation pertains to the fact that skilled L1-readers are very familiar with the written form of their native language, and automatically decode every grapheme (i.e. letter or cluster of letters) they read by producing a phonological representation of the sound (Snow,2002). This means that, when a learner reads a foreign language word its Working Memory will automatically match that sound with a first language phonological representation (i.e. will pronounce it the first language way). Thus, even if that learner reads a given word aloud following the teacher’s rendition of it, the L1 phonological representation of that word in the learner’s Working Memory will cause interference, with negative consequences for learning.

Another less recent finding (Neufeld, 1979) suggests that second language learners’ pronunciation might benefit from a mute period – a period of intense auditory exposure to L2 before attempting to produce the sounds. In Neufeld (1979) students were trained on pronunciation of sounds from Inuktitut, a language to which they had not been exposed previously. The learning process involved intensive listening to the language, with no attempt at producing the sounds. They were later instructed to produce the sounds and their attempts were rated as being mostly native-like. Neufeld claimed that the silent period at the beginning helped the students to accurately produce the language later. Removing students’ own attempts allowed perception to remain more plastic, such that the L2 acoustic template is heard accurately before erroneous phonetic utterances in L2 become fossilised. Producing the sound too early, and therefore incorrectly, would have influenced this acoustic template and thus hindered their production.

A mute period may prove beneficial in enabling the learner to hear (and thus produce) subtly different phonetic features, new phoneme distinctions and unfamiliar sequences of stress patterns. One possibility is that an artificially induced mute period may protect the learner from using first language phonological categories to represent the L2 system, thus enabling higher levels of production performance and avoiding L1 transfer or interference.

The threats posed by L1- language transfer to the correct uptake of  L2- pronunciation at the earlier stages of language acquisition are worrying only if we are aiming at 100 % accuracy due to the risk of fossilization, a phenomenon which, as explained in a previous post refers to the automatisation – often impervious to correction – of L2 learner errors. In most L2- classroom settings, unless we are training future international spies, we will be mostly aiming at clear and intelligible pronunciation with the majority of our learners and near-native accent only with a few talented ones.

How can we reduce the negative impact of L2-transfer on pronunciation?

Firstly, in order to avoid interference from a grapheme’s L1 phonological encoding (see the point made five paragraphs ago) on first introducing a new word it would be preferable not to expose the learners to its written form – this would avoid automated representation of the native phonological representation in Working Memory. In other words, it is better to present it orally, first in association with an image and, after some listening practice, to show it in its written form.

Secondly, L2 learners should be exposed to as much listening as possible in the context of a mute period before engaging in oral activities. Realistically speaking, in a typical state school classroom setting the pre-communicative mute period cannot be that long; but the most important lesson to be learnt from Neufeld’s (1979) research is that students should not be thrown into unstructured communicative practice straight away after presenting the target lexical items. The listening activities the students should be engaged in during this mute period should not only include test-like listening comprehension activities in the traditional sense: e.g. question and answer or true or false which focus solely on meaning. They should also include bottom-up processing activities that focus students on pronunciation and intonation, which involve matching words to sounds,  such as jigsaw listening, gap-fills with options to choose from or , at the basic level, circling a word or phrase from a choice of three or four options.

Thirdly, the observation that students seem to perform challenging L2 phonemes (sounds) more effectively when pronounced in isolation would seem to suggest – according to Simmonds, Wise and Leech (2011) that a babbling phase in which students imitate the target speech sounds in isolation might also improve non-native pronunciation. This can be done at the beginning of a lesson as a warm-up activity – I have done it quite often and it can be fun, depending on how you pitch it to the student and how you conduct it. Or, it can be set a homework activity to be carried out at home for a few minutes, recorded and sent to the teacher for feedback (were the  target sounds performed correctly? What could be done to improve them, etc.)

Finally, students need lots of practice in the context of structured and unstructured communicative activities. Such activities should, in my view, be staged after:

(a) effective modelling of the correct pronunciation;

(b) the mute listening period discussed above;

(c) extensive vocabulary practice through plenty of deep processing learning activities (e.g. the work-outs found on www.language-gym.com);

(d) Structured oral activities (e.g. find someone who; structured surveys; role-plays with prompts; timed oral translations) preceded by sufficient preparation time

(e) Less structured oral activities (at a later stage) in which students, through interviews, simulations, improvised role-plays, etc. converse freely about the topic-in-hand.

Traditional pronunciation drills (audiolingual style), minimal pairs and tongue-twisters or any other activities focusing students on pronunciation can be thrown in at the pre-communicative stage, provided that they maintain students motivation high and the students understand and accept the rationale behind them.

In conclusion, it is up to teachers to decide – with the course requirement they teach on as well as the interest of the stakeholders in mind, of course – how much emphasis they should put on accuracy. What research shows is that plunging students into unstructured oral communicative practice straight away is not beneficial to the development of accurate pronunciation. The above strategies may not always be easy or practical to implement but are in my experience very effective in enhancing student grasp and execution of the target language pronunciation.

The fundamentals of L2 vocabulary teaching

This article aims to answer the following questions:

  • What does ‘learning a word’ actually mean? When can we be satisfied that a student has actually learnt a given vocabulary item?
  • How can we enhance our students’ recall of the target vocabulary? How can we ensure that they do not forget what we taught one, five, ten and twenty lessons ago?
  • How can we effectively embed vocabulary instruction in the teaching of morphology and syntax? How can one ensure that vocabulary learning does not take over and that the whole lesson is not simply about learning to recognize or, at best, recall lexical items in isolation,  but also about deploying them through a range of functional and notional contexts in ways which are communicatively effective as well as morphologically and synctactically accurate?

1. What do we mean by ‘learning a word’?

1.1 Levels of vocabulary acquisition

Learning a word or lexical phrase involves more than memorizing its spelling, pronunciation and denotative meaning if one aims to use that word or phrase effectively in the real world (e.g. when grappling with an online article, attempting to understand a native talking to you in the streets of Paris or when writing an application letter overseas). Acquiring an L1 or L2 lexical item, may it be a word or a lexical phrase, also involves someone’s ability to master the following:

  • its morphology (e.g. if it is a French or Italian adjective, how is it affected by the gender and number of the subject?)
  • its word class (e.g. being aware that a word is a noun rather than a verb)
  • its other (denotative) meanings – in the case of polysemic words (e.g. ‘macchina’ in Italian means ‘machine’ but also ‘car’)
  •  any connotative meaning that word may have  (e.g. in English, is ‘chicken ‘ being used to refer to an animal, or is it used metaphorically to describe a coward?)
  • its collocation(s) (e.g. when learning the French for ‘go horse-riding’ one must be aware of the fact that in French ‘horse-riding’ is preceded by the verb “Faire”, to do, rather than “Aller”, to go).
  • how the meaning of a lexical item changes when in it is used in combination with other words (e.g as part of an idiomatic phrase)
  • its register, that is, knowing in which contexts it is appropriate or inappropriate to use a given word
  • any cultural ‘value added’ (e.g. knowing that a ‘cow’ in India is considered a sacred animal)

Hence, the practice of teaching words, as discrete items, divorced from any communicative and cultural context is not only limited but often flawed and misleading. This is less the case with denotative words such as ‘chair’ or ‘football’ than with words which are polysemic and/or loaded with connotative meaning(s). In what follows I shall focus first on how to maximize the recall of the basic aspects of vocabulary acquisitions, that is the memorization of the denotative meaning of a lexical item and of its spelling and pronunciation. In the second section of this paper I will concentrate on how to deal with the higher order levels of lexical learning.

1.2 Recognition vs Recall

Vocabulary acquisition goes through two stages. The first stage of acquisition is when the learner can recognize the word through its audio and/or visual representation. The second stage, involves being able to recall the lexical item and reproduce it verbally, either in its oral or written form.

Implications for the MFL classroom: in planning a lesson and defining the outcomes, decide which vocabulary items you intend the students to store in their mental lexicon as receptive vocabulary and which ones you want them to use actively and with what degree of accuracy, in their speaking and writing.

1.3 Level of accuracy and processing efficiency

Accuracy and speed of retrieval are two other very important dimensions of vocabulary acquisition. The faster an individual can retrieve the correct desired L2 word from Long Term memory, the more fluent and effective s/he will be in communicating the intended meaning. A vocabulary item that is, so to speak, ‘fully’ acquired, will be retrieved by the learner without hesitation with little cost in terms of Working Memory processing efficiency across a wide range of contexts. Obviously, recalling an item in isolation at relative high speed and with good accuracy is easier than doing that whilst you are holding a conversation across various topics. For learners, novices especially, it can be a very tall order.

Implications for the MFL classroom: make sure you include in your lessons/units of work plenty of opportunities for student to practise words in as many contexts as possible.

2. The fundamentals of vocabulary teaching.

2.1 Recycling and reviewing

Figure 1, below illustrates very clearly why recycling is important. The human rate of forgetting is such, that we already lose around 25% of what we attempt to commit to memory 30 minutes after having rehearsed it in Working Memory (henceforth WM). Seven days later, if no regular reinforcement has occurred, we will have lost 80 % of it. One month later, we will have forgotten virtually everything. That is why distributed practice is important and constitutes a more powerful way to consolidate memory than massed practice (i.e. better four sessions of 15 minutes a week every other day, than two sessions of thirty minutes two days away from one another ).

Implications for the classroom: Well, first of all, one should ensure that the words within a given lesson are recycled over and over again with several mini-check points every now and then to verify uptake and identify problem areas. Secondly, the students must be given plenty of opportunities to practise those words at home. Thirdly, words should be methodically recycled not just within the same unit, but across units – although this seems pretty obvious, I have rarely seen this happen in any of the institutions I have worked at: whenever one has completed a unit of work, the items taught in that unit should constantly and systematically be revisited in the context of every single unit of work to come. For instance, if we have just covered ‘staying healthy’ and we are moving to the topic of ‘holidays’, we could recycle some of the health-related vocabulary just learnt by discussing whether the food at the hotel the students were staying at was healthy and why; how the hotel’s menu could be made healthier; how healthy the students were during the holiday and what they are planning to do to get back into shape after two or three weeks of reckless eating and drinking, etc.

2.2 Factors facilitating recall

According to research, an exceptionally able student needs to have processed a word 4 times to learn it at its most basic level, an average student, 14-15 times. However, although the Latins used to say ‘Repetita juvant’, “repetitions help”, it is not simply how many times one comes across or repeats out aloud a vocabulary item which seems to be crucial in enhancing its recall. The following factors play a very important role in determining how efficient end effective memorization will be.

1. Shallow vs Deep processing – The more complex the cognitive operations involved in the learning process, i.e. the deeper the processing, the stronger the memory trace will be. On the shallow-to deep processing continuum we find, at one extreme, repeating word-lists aloud – the most classical example of shallow rehearsal. On the other end of the spectrum we find  problem solving activities where the brain has to think laterally and creatively (e.g. creating a complex mnemonic). Examples of  problem-solving activities commonly found in textbooks are sorting/categorizing activities, odd man out, riddles, definition games, etc. (www.language-gym.com has a great variety of these).

Implications for the MFL classroom: the foreign language teacher should try as much as possible to involve students in forms of deeper processing in order to speed up the learning process. This means going beyond the textbook page, as very few MFL textbooks designed for the British curriculum, provide sufficient recycling for the words they aim to teach (that is why I created : http://www.language-gym.com )

2. Spread of associations – Human forgetting is often cue-dependent; that is to say, the words may be in our Long Term Memory (henceforth LTM), but we have lost the ‘access’ code so to speak to get to them.  Research has clearly shown that  the greater the number of associations/connections that we create at the physical (e.g. graphemic, phonemic, etc.), semantic and emotional level with pre-existing material in LTM, the greater the chances will be for the target item to be retrieved successfully and efficiently in the future. The explanation for this is that when WM (Working Memory) is trying to fish out the word we need from LTM, all the words related to it in meaning, spelling and sound, and word class become activated automatically, especially those that are closer in meaning and end and start with the same letters.

Implications for the MFL classroom: teachers should include as many opportunities in their lessons for new vocabulary to be linked to previously learnt one so as to create as many connections as possible. The more elaborate (deeper) the connections, the better. Point 3, below, mentions other forms of associations which widen the range of possible connections we can make.

3. Synergy of stimuli – empirical evidence has shown (e.g. Paivio, 1981) that using different stimuli synergistically to appeal to various senses simultaneously may enhance recall. This is why a lot of us have used or still use flashcards. But this explains also why videos, by combining sound, images and often the spelling of the target words are even more powerful. Getting the students to respond to a video introducing new language items by emulating the movements they see on the screen would enhance the power of the video a notch further.

Implications for the MFL classroom – (1) On presenting words with denotative meaning for the first time try to use a video combining sound, picture and written form of the word; (2) When a word appears challenging, get the students to create a mnemonic which combines as many stimuli as possible. For instance, for the Italian word ‘occhiali’ (= glasses), one could picture in their mind a big pair of OCCHI (=eyes) with ALI (=wings) flying towards a pair of glasses and choose a suitable background music. I have used this technique personally a few times and has always been very effective.

4. Distinctivenes – distinctiveness refers to whatever makes the encoding (learning) of a given item in LTM ‘stand out’, ‘special’, more ‘vivid’. The factors making an item distinctive could be purely accidental (e.g.the teacher fell from a chair whilst teaching that item); intrinsic to that item (e.g. the target L2 word sounds funny, or like a swear words in one’s native language); there are personal, emotional circumstances surrounding the learning of that item that make it stand out (the teacher showed a picture whilst teaching that item, which evoked personal memories or triggered some strong emotions in the learner).

Implications for the MFL classroom: teachers should try to make the presentation of more challenging words as memorable as possible and/or teach the students to make them so, as they try to learn hem independently, by associating them with (a) powerful images; (b) items or situations in their lives which stir strong emotions, (c) humorous anecdotes etc. The way we pronounce words as we model their pronunciation can make a huge difference in terms of their distinctiveness, too. It is not rare for students to complain about how dull the voice of their teacher or of the actor on the recording is – surely, dullness is the antonym of distinctiveness.

5. Personal/affective investment – this refers to the processing of the to-be-learnt item that taps into our affective world, our own personal experiences related to it and its relevance to our lives.

Implications for the MFL classroom: include activities which involve a degree of personal response. for instance, when teaching adjectives, ask them to use them to describe their best friend, favourite cousin, pets, etc.

6. Target-item learnability – One dimension of a word’s learnability refers to the intrinsic challenges posed by the word to Working Memory. A word is more difficult to process and therefore learn when it is hard to pronounce (Baddeley, 2005); when it is so similar to an L1 item as to cause ‘cross-association’; when it is long (this is due to the fact that WM efficiency is quite limited as it can only process between 5 and 9 digits at any one time -Miller’s magic number). Other threats to learnability may intervene when the target items are not seen by the learners as relevant to their interests/goals and when their meaning is fuzzy, unclear. Moreover, generally, abstract words which are more connotative in meaning, tend to be less easily learnt. The word class the item falls into will also affect its learnability; for instance function words (e.g. prepositions, indirect object pronouns, etc.) are going to be less easy to be recalled as they are less semantically salient. Finally, the extent to which the words taught in a lesson are semantically related will affect their intrinsic learnability. 

Implications for the MFL classroom: (1) when selecting which vocabulary items to teach, consider the threats to learnability posed by the the first language of the student and devise some strategy to enhance their learnability using the tips above; (2) when more than one word exist in the target language for an item, choose the one that is more learnable, especially if it is more frequent than the others anyway; (3) You may want to teach the learners to break up longer and/or more challenging words into chunks in order to make it easier and more efficient for the articulatory loop in WM to process the item; when possible, break the word up into chunks which resemble words in the students’ first or second language.

7. Focal attention on the target item – although this factor is the most important, I left it for last because is the most obvious of them all: for any effective learning to occur, the students must be focused on the target stimulus. All of the above will be meaningless if the students are distracted, as interference during rehearsal is the most lethal cause of forgetting. The most important fact to note is that any given information does not last in WM for longer than 15-30 seconds without rehearsal; if any disruption to attention occurs and no further rehearsal of that item takes place, forgetting by interference will occurr.

Implications for the MFL classroom: obvious, but not easy: make your teaching as motivating, engaging and stimulating as possible.
(to be continued)

Five things I do when I correct my students’ essays

download (1)

My Ph.D study,Conti (2004), (as cited in Macaro,2004 and 2005, Ko Yin Sun, 2009, Goonshooly, 2012, Barjesteh, 2014, Cohen and Macaro, 2014, etc.) has provided me with great insight into the strengths and limitations of error correction. The following is a very concise list of what I believe to be the most important strategies to deploy in the error treatment of surface level errors in foreign language writing.

0. Caveat

Please note, this is something teachers can afford to do when they have a relatively light timetable or with specific students who are particularly problematic and need a lot of attention. I wouldn’t recommend this approach with every single class and student of yours as it is very time-consuming. In language instructions the focus should be on teaching not on fixing.

1.Focus on the most important issues

No point in focusing on every single error you find in your students’ writing when you are giving individual feedback on their essays. There is only so much attention a student can invest in the remedial learning process. Select only a few errors (3 to 5) at a time using the following criteria

  1. Errors that can be treated – no point in focusing an absolute beginner learner on mistakes involving the use of the pluperfect … Only treat errors for which the learner is developmentally ready;
  2. Errors that seriously impact understanding – these errors are the most important to deal with as they mislead the reader;
  3. Errors that keep recurring and seem impervious to correction – these errors need a lot of attention because once an error is fossilized it is very difficult to eradicate. Since these errors require a lot of work, try and prioritize the ones which, in your professional judgement, are more important (e.g. the ones that mighty penalize a student in a forthcoming exam);
  4. Errors that the learner would like to eradicate – it is my belief, controversial amongst some colleagues, that the learner should have a say as to what they should address in their remedial learning process. The rationale for this is that since s/he is going to be main the agent in this process, the fact that s/he chooses which errors to target may enhance their intentionality to eradicate error.

Do not address, in individual feedback, errors that are common to most of the class, as they can be the focus of a series of remedial lessons for the class as a whole

2. Find out what the root causes of error REALLY are

One common mistake teachers make in the corrective process, is to give all errors the same blanket treatment (be it direct/indirect correction with or without explanation or editing instruction) as if they were all caused by the same cognitive process(es). A bit like some doctors do, by giving a broad spectrum antibiotic for any kind of infection.

Errors can be caused by either (a) a declarative knowledge failure (the learner does not know the rule) or (b) a Procedural knowledge failure (the learner does know the rule and can self-correct, but did not apply it correctly or forgot to apply it in a given context because of processing inefficiency issues – e.g. cognitive overload, interference, etc.). It is important to identify the correct source of error before dismissing it as a ‘careless’ mistakes. There is usually more to an error than meets the eye.

In my study I used a number of research tools to investigate my subjects’ errors and the best one was definitely asking the students to edit the essays they wrote under think-aloud protocol conditions (i.e. they verbalized their thoughts as they attempted to correct). The knowledge I gained from that process was crucial to the success of my error treatment experiment.

3.‘Make it personal’

In my opinion, like any other type of instruction, error correction is greatly enhanced by making it as personal as possible a process, especially when we are dealing with weaker and/or less confident learners. One-to-one conferences are the best way to start the never-ending dialogue between teacher and student that the corrective process should really become. Using the page or the audio track as an interface between the student and the teacher makes the process much more distant and impersonal; the human contact, on the other hand, especially in the presence of judiciously gauged motivational feedback can do wonders for student’s self-efficacy and intentionality.

Let us not forget that the teacher’s role in the success of any remedial learning is crucial just as it is in any other kind of instruction. I often use the analogy of the person who wants to lose/gain weight in the gym. If you look at the rates of people who carry on training after the first three-four sessions, those with a personal trainer/life coach are less likely to drop out by a whopping 50 %! Why? Because a lot of us need encouragement, reminders, praise and, sometimes, a good telling-off…

When embarking on the remediation process, the teacher needs to take on a role alike the one of a ‘personal trainer’ since, as I shall explain below, errors are not eradicated in one go, it may take months or in certain cases, when an error is fossilized, even years. S/he will have to remind, prod, encourage, push the learner to keep working on the target mistakes.

It goes without saying that like every personal trainer the corrector must be inspiring and empathetic both emotionally and cognitively with the correctee.

4.Ensure there is a serious and sustained cognitive investment on the learner’s part

Several studies including mine have identified lack of student cognitive/personal investment in the error treatment as a major determinant of the failure of corrective interventions. Student writers do not look at the teacher’s corrective feedback and when they do they are superficial and do not follow it up. Teachers often do the same. They do a one-off remedial lesson on finding masses of students making the same mistake, then they move on. What I learnt in the course of my investigation is that for error to be eradicated (as mentioned above) both teachers and students must work hard. The students must put a lot of effort in the process at many levels: research, self-study, writing practice, self-monitoring and introspection.

Scaffolding the feedback-handling process in order to involve the student actively in the process is crucial, in this respect. Feedback-handling activities that students may be asked to perform on receiving feedback include: explaining the teacher correction; hypothesizing why the mistakes was made; describing what the rule that was broken is; producing student-generated examples of that rule across various contexts; produce a mini-lesson to deliver to a group of peers,ect.).

In my study, all of my informants reported drawing great benefits from such activities as they enhanced their self-knowledge as to the mistakes they were more likely to make to a point that they reported looking for those mistakes without much thinking before handing in their written pieces.

Another ingenious way of involving the students in the corrective process is to ask the students to step in before the essay is even completed and the feedback given. How? By asking them to annotate on margin whilst writing the essay any doubt they may have about the deployment of a grammar structure or lexical item. I use this technique a lot and it pays great dividends. This technique, that I call LIFT (Learner Initiated Feedback Technique) is dealt with in greater detail in a dedicated post of mine on this blog (here: ‘L.I.F.T. – an effective writing-proficiency and metacognition enhancer’).

5.Provide extensive practice

Many interventionist studies which involved editing instruction have failed whilst others have succeeded in enhancing grammar and/or lexical accuracy based on their duration and intensity. As already hinted above, learners need extensive practice to eradicate the target errors. Why? Because in learners’ Interlanguage system the wrong and correct representations of a grammar rule that has not been fully or correctly learnt coexist and often have equal weight (or, when the wrong form is fossilized, this will have greater weight). This entails that when the brain needs to apply that specific grammar structure the correct and the incorrect representation will both compete for retrieval. Extensive practice (highly monitored at the beginning) is required for the correct representation of the rule to acquire greater weight until it has become so strong in terms of memory trace to win the retrieval ‘competition’.

The extensive practice envisaged should occur:

  1. across a wide range of semantic contexts;
  2. in syntactically simple sentences to start with, moving gradually to more complex and longer chunks of text;
  3. in highly monitored performances (such as non-timed essays/translations) to start with and at a later stage, in the context of less monitored ones (such as timed essays or oral conversation).

Teachers are very busy people and one cannot always do all of the above as well as they would like to. However, these strategies can make a serious difference, in my personal experience, when applied to error treatment consistently. I suggest, if one does not have the time to do all of the above with every single student one teaches, to implement these strategies at least with the most needy of our learners, or with the ones that currently, in you opinion, are not gaining much benefit from your corrective feedback.

I deal with the issue of correction much more extensively in a research-based article of mine (‘Why teachers should not bother correcting errors in their students’ writing (not the traditional way at least’) : 

More on this topic in the book I co-authored with Steve Smith : ‘The Language Teacher Toolkit’ available on http://www.amazon.co.uk and http://www.amazon.com. 


Five common pitfalls of foreign language grammar instruction

Five common pitfalls of L2-grammar instruction

Most teachers nowadays agree that grammar instruction plays an important role in effective foreign language instruction. There are, however, a number of important factors that one has to take into consideration in the planning, delivery and evaluation of the effects of grammar instruction which are often overlooked, thereby undermining its efficacy. Such factors, which relate to both our epistemological assumptions about the nature of language learning and to neuroscience in general, can seriously undermine the effectiveness of our teaching as well as be important causes of daily teacher frustration. Here are five of the most common pitfalls of grammar instruction that I have witnessed in nearly three decades of MFL teaching.

Pitfall 1: too much focus on declarative knowledge 

A few years back I inherited from a colleague the ‘dream year 9 class’, a class that is, which included la crème de la crème of the students in my school. I was excited as I had taught some of these students before and I knew how keen and bright they were. On the first day of meeting them, one boy, Shaneel, told me ‘Sir, we have already learnt ALL of the (French) tenses!’. I was a bit skeptical but gave them the chance to show off their prodigious knowledge of the five tenses in question by asking them to translate sentences from English to French on mini-boards. The result was disastrous. ‘Wait, Sir’ this is not how we learnt it!’ protested Shaneel, we learnt them like this, and he recited to me  – fairly accurately – scores of verb conjugation tables much in the same way I had learnt Latin at Grammar School.

In other words, Shaneel had declarative knowledge of the Present, Perfect tense, etc. and all of the other grammar structures he had been taught, but was not able to transfer that knowledge to real life use (especially in the oral medium) across various semantic contexts. The reason? He had not been given enough opportunities in lessons, to acquire executive control over the target grammar structures across the following dimension of learning: (a) skills/modalities (L,R,S and W); (b) semantic areas; (c) communicative pressure.

The dichotomy Declarative/Procedural refers to the distinction between having intellectual knowledge about a target structure as opposed to the ability to apply the same knowledge subconsciously bypassing working memory’s attentional systems. According to many models of second language acquisition these two types of knowledge are completely separated, and several of them (see Stephen Krashen’s, for instance) posit that declarative knowledge can NEVER be converted into procedural knowledge. In other words, knowing a grammar rule, does not equate in the least with being able to use it spontaneously and automatically in unmonitored communication.

The obvious Implication for grammar instruction is that to assume that students have acquired a given structure based on their recall of grammar conjugation (by rote) or their effective executions of gap-fill activities is completely erroneous. The acquisition of a grammar structure takes a very long time and cuts across many dimension of morphology, syntax and meaning; hence, teachers should not feel as frustrated as in my experience often do at seeing structures that have been taught over and over again being deployed erroneously by their students in their oral or written output. It may simply mean that more extensive practice is required – not necessarily more intellectual knowledge.

In conclusion, online verb conjugation trainers (e.g. www.language-gym.com), Gap-fill exercises and all other activities aiming at enhancing morphological manipulation skills are useful but must be used in conjunction with translation and scores of real time communicative (oral and written) tasks.

Moreover, if we accept the notion advanced by most psycholinguists that intellectual knowledge about grammar (which is the one student obtain through correction and formative assessment) does not really impact acquisition, we will understand why error correction often has very little impact on our students’ mastery of the most complex structures (see the first article on this blog, below)


Pitfall 2: Developmental ‘unreadiness’

The Natural Order of Morpheme Acquisition Hypothesis is a theory based on a fairly large body of evidence which seems to indicate that humans acquire grammar in an order predetermined by nature. Although I do not espouse this theory of Language Acquisition, researchers working in this paradigm have gathered useful evidence indicating that there are some developmental constraints which limit our brain’s ability to learn the target language grammar structures. Such constraints are due to the challenges posed by such structures to the developing linguistic skills of the L1/L2 learner at given moments in time. To use an analogy: before teaching someone how to park, you would teach them how to start the engine, reverse, how to engage the clutch,etc. By the same token if grammar structure X requires the knowledge of grammar structure Y for its effective execution, one would have to be able to perform Y effectively before being able to learn X. The list below shows, for instance, the order of first language acquisition of English Morphemes in R. Brown (1973):

1 Present progressive (-ing)

2/3 in, on

4 Plural (-s)

5 Past irregular

6 Possessive (-’s)

7 Uncontractible copula (is, am, are)

8 Articles (a, the)

9 Past regular (-ed) 10 Third person singular (-s)

11 Third person irregular

12 Uncontractible auxiliary (is, am, are)

13 Contractible copula

14 Contractible auxiliary

Although I do not believe that the above order is necessarily correct and all of the evidence produced in its support valid, the Natural Order Hypothesis points to the importance of developmental readiness and has  one important implication for language learning: that without getting bogged down with which morpheme comes first or second or thirdwe need to sequence the order in which we teach grammatical structures  judiciously, based more on our cognitive empathy with the students and our experience of teaching equivalent groups of learners in the past rather than on the textbook or schemes of work provided by the Ministry of Education of Local authorities. Our presumptions of what constitutes an easy or challenging grammar structure for our students may not coincide with our student’s developmental readiness to acquire it. Grammar structures must be taught and corrected only when the students are developmentally ready to acquire them, in order for grammar instruction to be effective.


Pitfall 3: The rate of human forgetting / Poor recycling

Picture 1, below, shows the way us humans ‘forget’ the information we have been initially exposed to. As many studies have clearly proven, after only two days 70% of what we have been taught/processes on day 1, is lost. After seven days without any memory rehearsal, about 80 % of it is forgotten. Unless, through constant recycling, the modelling of effective revision strategies and continuous formative and summative mini-assessments teachers keep the memory traces alive, the human rate of forgetting is such that even the best grammar lesson will be forgotten.

Picture 1- Rate of human forgetting


Unfortunately, in my career I have rarely seen grammatical structures (or even vocabulary) being recycled constantly and in a principled way – fear of interference being often the obstacle, e.g. : if I revise the present tense now that I have just introduced the Perfect Tense, my students will be confused. Constant recycling, however, is absolutely imperative.

Poor recycling often also explains the inefficacy of a specific type of grammar instruction, correction, and why spending hours recording formative feedback on an App like Explaining everything may simply not impact learners much. The intellectual knowledge – which may not become procedural for the reason mentioned above – produced by these means will be mostly lost unless teachers provide constant recycling of that feedback over the five six weeks following the provision of that feedback (honestly: how many teachers actually do that?).

Pitfall 4: Low cognitive empathy

Cognitive empathy is a term I coined a couple of days ago during a discussion with a colleague. My point being that to be an effective teacher you must not simply be emotionally empathetic but also in sync with the learners’ thought processes and general cognitive development. AFL strategies do help a lot in terms of giving us an insight into how well our students our doing in learning what we are teaching them. However, they do not give us sufficient insight into the cognitive barriers to acquiring a specific target grammar structures. For instance, in the planning phase of teaching the Perfect tense in French one should think about all the possible obstacles posed by the following to the students cognitively (not just in terms of intellectual learning, but also in terms of acquisition as defined above):

  • The learners’ grammar background (how well do they master the present tense of AVOIR and ETRE?; do they know how to pronounce ‘e’ with an acute accent?; etc.)
  • The native language (does their native language have an equivalent of this tense? Will it cause interference?)
  • The various steps needed to be able to master the perfect tense (deciding whether it is the correct context for Perfect Tense use, correctly choosing the required form of the correct auxiliary; deciding if the verb is regular or irregular, select the correct regular or irregular form of the past participle; pronouncing it correctly)
  • The time available for it to be learnt DECLARATIVELY;
  • The time available for it to be learnt PROCEDURALLY (i.e. automatized)

Another useful strategy to deploy whilst planning our lessons is to cast our mind back to the days when we learnt the same grammar structures as L2 learners of French: what did we find hard? What strategies did we come up with to facilitate our own learning? How long did it take us to learn that tense? Our (L1 English) students will have more or less the same issues, after all. This should enhance our cognitive empathy.

Finally, there are useful techniques that I have used several times to gain a better insight into our students’ learners cognitive processes. One of them is think-aloud protocols a very powerful (if time consuming process) tool to get into our students’ minds:  students perform a task (e.g. writing an essay) whilst verbalizing every single thought that goes through their heads. Getting them to write an account of a past holiday after a cycle of lessons on the Perfect tense in front of you while thinking aloud will provide you with a clearer understanding of how well they master the Perfect Tense in real operating conditions and of their problems with that tense (http://en.wikipedia.org/wiki/Think_aloud_protocol )

I have observed poor cognitive empathy in many lessons over the years. Students resent it as much as they resent lack of emotional empathy or humour. High levels of cognitive empathy are, in my opinion, the marker of an excellent practitioner.

Pitfall 5: Lack of a common metalanguage

If teacher and students do not share a common metalanguage, grammar instruction is less effective. Several studies have shown that students who do have a solid repertoire of metawords (adjectives, mood, tense, ect.) learn grammar more effectively both in and outside the classroom (independently). They also learn more effectively from corrective feedback. One anecdote I will never forget was when I inherited a class from my former colleague Gill Bruce and since she had taught her students the difference between an adverb and an adjective, I could –for the first time ever – very quickly get my students to understand the difference between ‘mal’ et ‘mauvais’ in French by simply saying: one is an adverb and the other one is an adjective.

The implications for teachers is that we need to use metalanguage from the very start and make constant reference to it (in our marking, too).

In conclusion, in planning and delivering our grammar lessons one has to be very mindful of all of the above factors. We are often reminded by scholars and educators of the importance to empathize with our students emotionally since, as my colleague Dr Michael Browning rightly said to me once: “if they like you a lot they will learn better anyway, regardless of which technique or technology you use”. However, empathizing cognitively in terms of truly attempting to (a) sync our teaching to their specific individual linguistic needs and (b) to the way their brain works when acquiring a foreign language (as posited by neuroscience) is imperative in order for us to pace our teaching effectively and to intervene effectively through adequate remedial instruction.

Why teachers should not bother correcting errors in their students’ writing (not the traditional way at least)

Metacognitive enhancement and error correction: a discussion of the shortcomings of traditional error treatment and of the potential benefits of learner training in self-monitoring strategies

Dr Gianfranco Conti

Garden International School (Kuala Lumpur)

Dr Yoko Sato

Hosei University (Tokyo)


This paper evaluates the merits of the most traditional forms of error treatment, Direct and Indirect Error Correction, as applied to L2-writing, highlighting the epistemological, pedagogical and learner-strategy related issues which undermine their efficacy. It also discusses why Explicit Strategy Training in self-monitoring strategies may constitute a more promising alternative in the light of Cognitive theories of L2-learning and of the existing empirical evidence.


Research into the impact of teacher corrective response to L2 writers’ sentence level errors indicates that the most popular approaches, Direct and Indirect Correction (i.e. learner errors are either corrected by the teacher, or marked and left for the learners to self-correct) do not have a significantly enhancing effect on L2 writing accuracy and writing proficiency development (Truscott, 1996; Macaro, 2004). As Macaro (2004: 242) rightly observed,

these results are somewhat dispiriting given the amount of effort

that teachers put into providing feedback to written work. …

the evidence does seem sufficiently strong to question the effort


To make matters worse, other strands of Error Correction (EC) research have identified a number of problems which seriously undermine the efficacy of these techniques. Firstly, teacher feedback is often unhelpful and misleading (Hedgcock and Lefkowitz, 1998). Secondly, learners rarely pay attention to teacher corrections and, when they do, they do it superficially and/or applying ineffective feedback-handling strategies (Cohen, 1987; Cohen and Cavalcanti, 1990). Thirdly, too much correction may engender learner anxiety and error avoidance behaviour, both potential inhibitors of L2 learning (James, 1998). Furthermore, there is no consensus as to which feedback techniques are more effective in which context, and whether feedback on surface level error should be administered at the beginning or at the end of the multi-draft process currently so popular in L2 writing instruction.

The reactions to this state of affairs have ranged from proposing a total ban on correction (Truscott, 1996) to exploring innovative approaches which integrate traditional corrective approaches with other error-treatment techniques (Macaro, 2004). In particular, as we shall argue later, a very promising way forward for error treatment methodology seems to be indicated by a handful of Explicit Strategy Training (EST) studies in Self-Monitoring strategies (Ferris, 1995; Conti, 2001; Macaro, 2001). Below, we evaluate the merits of traditional corrective approaches and those of EST by reviewing what we consider to be the most representative empirical studies published in the core literature (see Ferris, 2004 for a fully comprehensive review).

2 Studies investigating the impact of Direct and Indirect Correction (with or without supplemental focus on form instruction) on L2 writing accuracy

2.1 Different forms of EC compared

Semke (1984) investigated the effects of EC on the writing of 114 American university students of L2-German. She divided her sample into four groups, each receiving one of the following types of teacher correction on their journal entries:

1. Comments on content only

2. Comments on errors only

3. Comments on both contents and errors

4. Errors underlined; learner self-correction expected

Using error-free T-units to measure accuracy, Semke found no significant differences among the groups in terms of error reduction. In addition, Group 1 (comments on content only) wrote longer freewrites at post-test and was significantly better than all the others on fluency and on a cloze test. This was interpreted as indicating that learner progress can be enhanced by practice alone and that EC could not just be unhelpful but even harmful. In fact, Semke reported that corrections may affect learner attitude to writing negatively, especially when they are required to make corrections by themselves.

Similarly, Kepner (1991) investigated the impact of different types of feedback on the journal writing of 60 second year university students of L2-Spanish. The sample was divided into two groups, one receiving DC with explanations and the other comments on the content only . The study covered a term and consisted of six pieces of work. Kepner found that the latter group produced significantly more higher order proposition than the former group, with no statistically significant difference in the number of errors made between the two groups.

Robb et al. (1986) studied the effects of EC on the writing of 134 Japanese university EFL learners randomly assigned to the following groups, each receiving a different form of error treatment:

1. Direct correction, indicating the errors and the correct forms;

2. Correction code to point out type and location of errors;

3. Highlighting the locations of errors without any explanation;

4. Marginal tally of the number of errors for each line, with no cueing as to the

typology or location of the errors

Students were required to redraft their compositions, making the appropriate changes. Writing ability was assessed using 19 objective measures, three of which was used to measure accuracy: the ratio of error-free T-units to the total number of T-units, the ratio of error-free T-units to the total number of words and the ratio of words in error-free T-units to the total number of words. At the end of the course, the researchers found no significant differences in students’ writing ability. Thus, they argued that less time-consuming methods of directing students’ attention to surface error may suffice (Robb et al., 1986: 91).

2.2 The effects of DC with supplemental editing instruction

Polio et al. (1998) set out to investigate whether additional editing instruction – the innovative feature of the study – would enhance learners’ ability to reduce errors in revised essays. 65 learners on a university EAP course were randomly assigned to an experimental and a control group who wrote four journal entries each week for seven weeks. Whereas the control group did not receive any feedback, the experimental group was involved in (1) grammar review and editing exercises and (2) revision of the journal entries, both of which were followed by teacher corrective feedback. On each pre- and post-tests, the learners wrote a 30-minute composition which they were asked to improve in 60 minutes two days later. Linguistic accuracy was calculated as a ratio of error-free T-units to the total number of T-units in the composition. The results suggested that the experimental group did not outperform the control group. The researchers conjectured that the validity of their results might have been undermined by the assessment measure used (T-units) and/or the relatively short duration of the treatment. They also hypothesised that the instruction the control group received might have been so effective that the additional practice for the experimental group did not make any difference.

2.3 The impact of selective DC

Other studies had a much narrower focus as they investigated the effects of EC on only a few selected structures. Loewen (1998), for instance, conducted a small-scale investigation of the effects of selective EC on the writing of twelve mixed nationality ESL learners. Loewen divided the learners into two groups, one focusing on the Simple Past and Plural s, and the other on the third person –s and the passive voice. The treatment was provided in the form of Direct Corrections on the seven essays written by both groups. At the end of the semester no significant improvements were observed for either group. Loewen acknowledged that the small sample size and the absence of control groups were important limitations of the study. What might have also affected the results was the infrequent occurrence of three of the four target structures, partly due to the learners’ avoidance, and partly due to the type of expository writing the students engaged in, which did not present many obligatory occasions for the use of those structures.

2.4 The effect of IC on re-drafts

Another set of studies investigated the short-term effects of IC on the accuracy of student rewrites (Fathman and Whalley, 1990; Ashwell, 2000; Chandler, 2003); the students were corrected on their writing and then asked to rewrite the same paper, taking the corrections into consideration. These studies indicate that even when the students are solely focused on form and are told where the mistakes are, they can only correct from 1/3 to1/2 of the errors. The success rate increases to 90% when they are given the actual rule and need only copy – which means that even in the best conditions they still miss 10% of the errors. Not unsurprisingly, Krashen (2005) concludes his review of these studies stating that their findings constitute ‘hardly a compelling case for correction’. Finally, it should be noted that, none of these studies being longitudinal, we do not know whether the error treatment had any long-term impact on the participants’ proficiency development.

3 Why Traditional Approaches to Error Correction do not work

In drawing conclusions from the existing research, one has to bear in mind that EC studies differ greatly across a number of important design and procedural features, such as corrective techniques, learner types, languages, time scale and assessment measures. This makes the comparison across studies difficult thereby impinging on the generalisability of the findings (Ferris, 2004). Moreover, there are major flaws in design and methodology that need to be taken into account. Firstly, the measures of accuracy adopted by many studies, error-free T-units based counts, are not the most effective instruments for detecting error reduction, especially over relatively short periods of time and at lower levels of proficiency. For instance, even if a student goes from producing an average of ten errors per T-unit at pre-test to producing only one error at post-test, the Error-Free T-unit count will not register any improvements whatsoever.

Secondly, the number of written pieces the past studies involved was sometimes far too small to ensure sufficient practice of the problematic items and exposure to the relevant corrective feedback (e.g. in Kepner, 1991) .

Thirdly, several studies were either not transparent as to the inter-rater reliability procedures adopted to validate the error counts (if any) or did not publish the figures (Ferris, 2004). For instance, Robb et al (1986) stated that their reliability score was .87; however, they did not tell us to which of the 19 objective measures they used that score refers. It would appear that the score was the average of all the 19 measures (which included objective measure like number of words, number of clauses and T-units). Thus, we do not know the reliability of the actual coding of the accuracy measures used. The fact is that it is undoubtedly easier to get a high reliability on measures like number of words, number of clauses and T-units; it remains to see if, taken in isolation, the measures-of-accuracy inter-reliability scores would have, by themselves, been sufficiently significant .

Furthermore, very few studies provided information as to the learning background and beliefs of their participants, which may play crucial roles in the success of a corrective approach. A student who has little metalinguistic knowledge may find grammar feedback daunting. On the other hand, someone who feels that Direct Correction with grammar explanations has worked very well for him/her in the past may resist other corrective treatments, especially if he/she knows that they are only temporary measures for an experiment. It should also be noted that when one tests an instructional technique, one is also testing the effectiveness of the teacher who delivered it, his/her rapport with the students (with its enormous implications on motivation) among other things. None of the published EC studies reported on this crucial element of learning. In fact, in reviewing the above research, one cannot help notice the absence of qualitative data about the human factors that may have interacted favourably and unfavourably with the treatment under investigation.

In view of the above limitations, the existing empirical evidence cannot be said to conclusively prove that EC does not work. It does, however, strongly indicate that Direct Correction (with or without rule explanation) and Indirect Correction do not significantly improve L2-writer accuracy. The finding that the combination of Direct Correction (DC) with grammar instruction and peer editing may not work is particularly important given that a host of EFL books and materials promote and/or provide practice in this kind of approach (Ferris, 1999). It is also noteworthy that Indirect Correction (IC) may induce some degree of learner disaffection (Robb et al, 1986).

A number of factors undermine the effectiveness of EC, the most important ones being the following:

1) Learners’ attitude to correction is often superficial and their feedback-handling strategies are poor.

The effectiveness of DC depends largely on two factors: (a) how effective the corrections are in helping the learners restructure their assumptions about a given L2 item and (b) what the learners do with the feedback. As Truscott (1996) noted, numerous studies have revealed that the quality of teacher feedback is often poor. Furthermore, there are findings that learners tend to have a superficial attitude to correction and invest very little cognitive effort in handling teacher feedback, often simply looking briefly at the mistakes and making a mental note of them (Cohen, 1987; Cohen and Cavalcanti, 1990; Conti, 2001). IC does partly address this issue by involving the learners more actively in the feedback-handling and editing process through self-correction. However, learners may not be able to correct (Krashen, 2005), may correct wrongly (Conti, 2001), or may be frustrated if asked to self-correct what is well beyond their current grasp of the language. In addition, accurate self-correction could be the results of applying the ‘guessing’ strategy’ “if a given item can only be X or Y and X in this case is wrong, then the correct form must be Y” (Conti, 2005). In such cases the students’ accurate self-correction may fool the teacher into believing that the learners have learnt the target item. More importantly, IC falls short of promoting learning of new structures and of re-learning of existing Interlanguage structures in that it concerns itself mostly with self-correctable mistakes. Thus, it leads at best to raising learners’ awareness of their mistakes and consolidating old material.

The role of the quality of teacher corrections and of students’ feedback-handling is crucial to L2 acquisition. The negative evidence provided by the teacher is often (especially in an FL environment characterised by low L2 input) the only way for students to realise that their usage of a given L2 item is wrong and needs restructuring. This process of “noticing” (Schmidt, 1990) marks the beginning of conscious L2 learning which, according to Skill-Theory , will eventually lead to the acquisition of the noticed item after extensive practice (e.g. Anderson, 2000; Johnson, 1996). However, this can only occur if the teacher’s corrections are clear and learnable (i.e. not beyond the learner’s grasp or level of acquisition) and if the learners pay sufficient attention to and apply any productive feedback-handling strategies (e.g. make a written note of the correction, ask an L2 expert for clarification, do follow-up remedial grammar work). Any form of instruction in order to be effective require the learner to take an active role and deep processing of the target information must take place. As Cohen (1987), Cohen and Cavalcanti (1990) and Conti (2004)’s study have shown, students’ cognitive investment in the corrective process is low and teachers do not seem to demand or even encourage it.

(2) DC and IC on their own are short-term measures which do not involve the learners in a sustained long-term effort to eradicate error.

This is problematic for two reasons. Firstly, the rate of human forgetting is such that after one week only 80% of whatever the students learn from the corrections would be lost without reinforcement. Thus, without some form of instructional follow-up, the impact of correction is likely to be minimal. Secondly, it should be pointed out that for any language items to be successfully learned, they must be practised with success (i.e. with positive feedback from the environment) on numerous occasions over a long period of time (Anderson, 2000). This implies that a one-off follow-up remedial lesson to address learners’ mistakes has very limited long-term learning potential. After all, two factors are crucial in any kind of learning: frequency of exposure and the amount of attention devoted to the target item/s; thus, for a correction to be learnt, any learners should be exposed to it frequently enough to first notice it, then learn it declaratively, and, finally, proceduralise it. In the case of fossilized items, in particular, the issue is complicated by the fact that de-fossilization is a very lengthy process. Furthermore, it is crucial that learners’ attention is constantly focused on the target item. This may not happen if learners are corrected on a large number of items over a relatively short period of time, and students are not asked or do not decide to prioritise certain items over others. Also, as Loewen (1998) pointed out, this will not happen if the structures targeted by the treatment, are not likely to recur often enough in the type of expository writing the participants are engaging in.

(3) DC and IC do not usually systematically attempt to identify the causes of errors, and tend to offer the same blanket-treatment to all mistakes and to all students.

It is one of the most fundamental assumptions of modern pedagogy that the success of any instructional programme depends largely on the correct identification of the target learners’ needs (Littlewood, 1984). Applied to the area of remedial instruction, this principle translates as follows; error treatment is likely to be more effective if teachers identify accurately the root causes of their learners’ mistakes. For instance, it would not make much sense for an L2 writing instructor to treat self-correctable mistakes caused by processing inefficiency (performance errors) in the same way as he/she treats errors due to lack of knowledge. The former may require training in editing strategies (see below) and a lot of practice; the latter, cognitive restructuring or, if they refer to grammar structure well beyond the learner’s developmental stage, possibly no treatment at all. The possibility that the “failure” of the above reviewed studies may be due to this issue cannot entirely be discounted.

(4) DC and IC rarely attempt to raise learner intentionality and motivation to eradicate error.

In the way DC and IC are usually administered, teachers do not systematically attempt to raise the learner’s intentionality to eradicate their errors. This is an important shortcoming since learner intentionality appears to play an important role in the success of any kind of instruction (Schmidt, 1994). For EC to work, learners must want to eradicate errors. As Johnson (1996) observed, the issue of intentionality is particularly important in the area of mistakes which do not impede communication and which learners may consequently not perceive as important – the so-called ‘minor’ mistakes. Learner perceptions of the importance and gravity of mistakes will depend on a number of factors. One such factor refers to the relevance to their academic success. If, as in the case of UK examination boards, a student can obtain a grade ‘A’ at GCSE having made a relatively high number of such minor mistakes, he/she will not put much effort on eliminating these mistakes. Another important factor will be the bias in corrective feedback consciously or subconsciously conveyed by the teacher to their students. If a teacher applies selective EC which prioritises effective communication over grammatical correctness, the learners may perceive the former as more important than the latter with consequences for the orientation of their intentionality.

Similarly, DC and IC may impact negatively on the motivation of those learners who do not know how to improve on their mistakes. In these writers’ professional experience, the perception of not being able to address one’s learning problems is a common and powerful catalyst of learner anxiety. Thus, whereas in highly self-efficacious learners correction may spark a productive form of arousal and/or facilitative anxiety (James, 1998), in less self-efficacious ones the opposite is often true. A counter-productive form of anxiety sets in and the students often linger in a state of learning inertia and disaffection. This issue is exacerbated when the correction on those problems is recurrent, judgmental and carried out without the due tact and sensitivity (Edge, 1989). Simply reminding learners of their inadequacies or asking them to self-correct what is beyond their ability could only be demotivating and detrimental to learning. As the resultative hypothesis (Skehan, 1989) posits, learners need to feel empowered with effective instruments for overcoming their problems and must experience success in their performance in order to gain confidence and feel motivated.

(6) EC is not traditionally a learner-centred process and does not aim at learner autonomy.

The emphasis on learner–centredness and autonomy is one of the most innovative features of modern L2 pedagogy. Yet, EC is still largely carried out through models of feedback provision whereby the teacher is the exclusive manager of the corrective process. The fact that teachers select the errors to be treated may have negative consequences on the motivation of those learners for whom self-determination is an important personal value. For such learners, an approach whereby the teacher allows the student to make an informed choice as to which errors they want to address may be more beneficial. Another issue with teacher-centredness and lack of learner autonomy relates to self-correctable errors often left in learners’ writing due to poor editing and carelessness. A number of studies have shown that many of the errors in L2 writing are self-correctable if their presence is pointed out by the teacher (e.g. Conti, 2004). As Ferris (1999) noted, L2 students must learn to become self-reliant and effective editors in preparation for their academic and/or professional life when teacher help will no longer be available. Eliminating self-correctable errors also has the practical advantage of allowing the teachers to devote their time to errors which the learners cannot correct by themselves. However, traditional EC does not equip learners with effective editing skills or instruct them to apply more attention and commitment on the editing process.

(7) Traditionally, Error Correction does not explicitly promote the enhancement of error-related metacognition.

Related to the previous point is EC’s failure to systematically develop learners’ awareness of their problematic areas. The effects of heightened learner self-awareness are manifold; the most obvious is that it may prevent specific errors from re-occurring by sparking a process of self-monitoring. This, in turn, may lead to more effective editing (e.g. “If I know that I make a mistake on this item, I will be more careful in using it in the future”) and/or independent learning (“Since I keep making mistakes on this item, I’d better learn more about it”). It may also trigger intentionality by bringing into the learner’s focal awareness the full extent of his/her lack of commitment and/or gaps in his/her mastery of the items (e.g. “So many silly mistakes! I must do something about it!”). Conti (2004) found that student -generated error checklist which included familiar mistakes noted in previous essay played a significant role in reducing self-correctable routinised mistakes due to processing inefficiency, especially those related to function words. These errors, as Johnson (1996) notes, are the most impervious to correction as they usually refer to less semantically salient items which often go unmonitored (Johnson, 1996) and unless brought into learner focal attention will keep going unmonitored until they are fossilized (Conti, 2004).

4 The way forward: Explicit Strategy Training in self-monitoring as an effective alternative

Should we ban EC altogether then, as Truscott (1996) advocated? This has been a thorny issue for the last decade or so and no-one as yet has refuted Truscott’s arguments convincingly. Surveys of student opinion clearly show that students want their errors to be corrected (Ferris, 2004). Authoritative Cognitive theories of L2 acquisition posit the crucial importance of EC as a catalyst of L2-learning (e.g. Byalistock, 1980; McLaughlin, 1987; Anderson, 2000); and there is some research evidence, however scant and controversial, that it can indeed work (Ferris, 2004). It seems that the real issue is not whether to abolish EC or not, but how to find a more effective alternative. As already stated at the beginning of this paper, alternative models of error treatment can already be found in the literature, in a handful of successful Explicit Strategy Training (EST). It should be noted that the remedial approaches adopted in these studies did not do away with EC, but rather integrated it with strategic instruction.

4.1 Rationale for the adoption of Explicit Strategy Training

The rationale behind the adoption of EST is that many of the factors which undermine the effects of EC may be viewed as strategic deficits, more specifically:

(1) the inadequacy of student interaction with teacher-correction maybe viewed as referring to a lack of effective feedback-handling strategies;

(2) the fact that students do not engage in long-term targeting of their errors, as stemming from a deficit in Self-Monitoring strategies;

(3) the recurrence of self-correctable ‘sloppy’ mistakes, as resulting from poor editing strategies;

(4) the lack of learner autonomy in the area of error-treatment, as caused by low levels of self-regulation (i.e. metacognitive strategies);

(5) the lack of declarative knowledge which is at the root of many errors may also be seen as the result of ineffective learning skills;

(6) finally, the disaffection of those learners who feel unable to improve may be seen as relating to the domain of affective strategies (e.g. resilience )

Viewed from this angle, remedial instruction which effectively addresses the above strategic deficits (e.g. Explicit Learner Training), should, at least in theory, have a more beneficial impact on learning than DC and IC.

4.2 Approaches to EST

The existing EST studies share the following features, which were lacking in the traditional forms of EC reviewed above:

(a) Enhancement of learner error-related metacognition

(b) A long-term process of self-monitoring

(c) Modelling of and extensive practice in the use of effective self-correction / editing strategies

(d) Personalisaton of error treatment

(e) Focus on the process rather than the product of writing and learning in general

(f) Synergistic use of various forms of EC

EST programmes typically consist of the following phases (O’Malley and Chamot, 1990; Cohen, 1998; Macaro, 2001):

(1) Pre-test needs assessment: The learners’ needs are assessed, usually through a combination of different instruments (e.g. questionnaires, interviews, think-aloud protocols and other forms of self-reports) in order to strengthen the validity of the data.

(2) Introductory phase: The rationale for the training is given and the target strategies are presented and modelled.

(3) Scaffolding phase: The learners receive extensive practice in the target strategies with the help of “scaffolding”, i.e., activities and materials which remind and encourage the learners to apply the target strategies.

(4) Autonomous phase: The learners are left to their own devices without any intervention on the part of the teacher.

(5) Evaluative phase: The learners’ use of the target strategies and their impact on their performance are verified. Normally the same diagnostic instruments used at pre-test are re-deployed here.

The rationale for adopting EST as an alternative to EC and IC lies in the fact that it addresses the deficits of traditional corrective methodologies that we have identified thus far, namely: (1) it is a learner-centred instructional methodology which has at its core the assessment of learner needs ( i.e. the analysis of the sources of errors in the context of error treatment). These enable the teacher to deliver more personalised, fine-grained instruction for different learners and different types of errors. (2) EST programmes usually include components for attitudinal change and for enhancing self-efficacy and motivation as well as metacognitive awareness (Wenden, 1987; 1991). These would in turn help to develop the learners’ self-regulation and intentionality to act on their learning deficits. (3) EST addresses the strategic deficits which undermine the effects of traditional EC (note 2). (4) being based on Skill Theory, which postulates that acquisition occurs only after extensive practice and numerous instances of feedback from the environment (Anderson, 2000; Johnson, 1996), EST provides/scaffolds extensive contextualised practice of the target strategies through tailor made resources/activities (Macaro, 2001).

Moreover, although empirical research in EST is still scant, some promising results have been yielded in a small number of existing studies in the area of L2 writing.

Lalande’s (1982) study, although the researcher did not refer to the study as an EST experiment, is considered as the oldest of the studies in this area since it involves metacognitive enhancement and strategic instruction. Lalande compared the effects of two different types of feedback on the writing of FL German learners: DC and IC. Upon reception of the marked essays the learners were asked to correct their mistakes and re-write the entire essay. For the experimental group, this involved interpreting the codes. As the course progressed, the experimental group learners monitored the frequency and recurrence of error types by referring to Error Awareness Sheets (error charts in which students logged their mistakes). Lalande found that “the combination of error awareness and problem solving techniques had a significant beneficial effect on the development of writing skills” and “effectively prevented students from making more grammatical and orthographic errors.” (Lalande, 1982: 78)

Ferris (1995) set out to improve 30 ESL writers’ essay accuracy through EST in editing skills which involves a strong metacognitive component (awareness-raising and self-monitoring). The corrective treatment included the following: self-correction, peer feedback, teacher- and learner-generated editing checklists, learners’ error chart, awareness-raising, remedial grammar instruction and exercises, and classroom editing activities. The essays were analysed in terms of five categories of errors: noun, verb, sentence structure, punctuation, miscellaneous. The results showed that nearly all her learners (28/30) made significant progress in reducing the overall percentage of errors in at least some of the five error-categories. However, they were not always successful in reducing their percentages of error in specific categories on which they had been advised to focus. Also, there was considerable variation in error reduction in the targeted categories across both error category and writing context (home or class). These mixed results, Ferris concluded, “show how individual a process editing really is” and that “personalized instruction and guidance in editing may be most effective” (1995a: 52).

Conti (2001) conducted a quasi experimental study involving 20 university students of L2 Italian of intermediate writing proficiency. He compared the accuracy of 250 essays written by two groups (about 10 per participant): one who received EST in feedback handling, error targeting and production monitoring strategies, and the other who received traditional DC. The results showed that the experimental group’s error rate decreased significantly from the first to last essay across all the categories of errors investigated. A comparison with the control group indicated a significant advantage for the independent variable. In fact, increases in error rate were observed for several comparison group learners. Analysis of think-aloud protocols collected during the writing of the last in-class essay revealed that most of the corrections by the experimental group students were through the application of the target strategies, as agreed by two independent inter-coders). The final questionnaires and interviews indicated that the experimental group felt their accuracy had improved substantially due to the training whereas the other group felt their accuracy had either improved slightly or not at all. It is noteworthy that there was no difference in the quality of the content of essays by the two groups as assessed by independent raters.

Macaro (2001) implemented an EST aimed at enhancing the writing skills of three groups of secondary school L2 French students. The training focused on composition and editing strategies using scaffolding techniques similar to Conti (2001). A set of memory, cognitive and metacognitive strategies were modelled and scaffolded over a semester in order to improve learner accuracy in the areas of tenses use, and noun-adjective agreement. The performance of three experimental groups was compared with three control groups who received DC. The results showed a significant advantage for the experimental groups.

It should be noted that with the exception of Macaro (2001), the existing EST studies investigated relatively small samples of students and each study investigated different foreign languages. This obviously undermines the generalisability of the findings. In addition, the comparability of the experimental and control/comparison groups had hardly been established statistically. Furthermore, apart from Conti (2001), none of the above studies calculated inter-rater reliability in analysing errors and strategy use, and this further threatens the validity and reliability of the findings. Nevertheless, the above review suggests the potential of EST as a more effective alternative to traditional EC in treating errors in L2 writing.

5 Conclusions

This paper has attempted to show why EST has the potential to be a more effective alternative to traditional EC. As discussed above, the existing research evidence strongly suggests that traditional EC does not work due to a number of intrinsic and extrinsic flaws. The intrinsic ones related to the epistemological assumptions that learning occurs by simply looking at the teacher correction; the extrinsic ones to the way teachers provide feedback or to the learners’ often superficial interaction with the correction. There are also other reasons for the failure of traditional EC not previously discussed in the literature: 1) lack of the identification of root causes of errors, whether cognitive or strategic, and of the treatment of errors accordingly, 2) insufficient provision of practice of and exposure to problematic items, and 3) lack of active development of learners’ motivation, intentionality and editing skills needed to become effective, self-autonomous writers. On the other hand, EST addresses these deficits. It 1) assesses learners’ deficits that lead to error-making or inadequate feedback-handling, 2) raises learners’ awareness of problematic areas and engages learners in a long-term process of self-monitoring of recurrent errors, 3) provides instruction and extensive practice in strategies that suit learners’ needs and have the potential, if used effectively, to enhance their performance and 4) gradually shifts the responsibility of the error-treatment process to the learners, who become self-reliant writers.

In the absence of a solid body of research, it is difficult to say with certainty that EST will definitely work and which form of EST is effective with what kind of learners. More studies should be carried out testing out the efficacy of the EST instruction with larger samples of students in a variety of contexts using more valid research methods than the existing ones. This state of affairs may deter teachers from adopting it in their classroom practice, especially when EST programmes are not easy to set up and implement as they require a lot more time, effort and know-how than simply correcting or coding learners’ errors. Furthermore, teachers must be trained in the use of needs assessment tools such as questionnaires, interviews and error analysis, and students need to be persuaded that the time and effort they are going to invest in the training are conducive to learning. Nevertheless, EST is worth the effort if one considers its long-term benefits. In fact, after the initial stages of the training where the learners require substantial cognitive and affective support, the students will become more self-reliant, and the teacher’s work will be limited to helping the learners solely with the errors that they cannot cope with. As the students mature linguistically and metacognitively year after year, the need for the teacher to intervene will be reduced even further. To start with, we suggest teachers incorporate in their practice the most important features of EST: 1) personalised error treatment, 2) enhancing the learning outcome of student-interaction with teacher feedback and 3) long-term, repeated instruction on problematic language items.


Anderson, J.R. 2000. Cognitive psychology and its implications (5th edition). New York: Worth Publishing.

Cohen, A.D. 1987. Student processing of feedback on their compositions. In A.L. Wenden and J.

Rubin. (eds.) Learner strategies in language learning. Englewood Cliffs, NJ: Prentice Hall, 57-69.

Cohen, A.1998. Strategies in learning and using a second language. Harlow: Longman.

Cohen, A.D. and Cavalcanti, M. 1990. Feedback on composition: teacher and student verbal report. In B.

Kroll. (ed.) Second language writing: Research insights for the classroom. New York: Cambridge University Press, 155- 177.

Conti, G. (2001) E possibile migliorare I’efficacia della correzione attraverso 1’istruzione strategica? –

Tuttitalia, 23, 4,14

Conti, G. (2004). Metacognitive enhancement and error correction: An investigation in the impact of self monitoring strategies on L2 Italian student writing. Unpublished Doctoral Dissertation, University of Reading, United Kingdom.

Edge, J. 1989. Mistakes and correction. London: Longman.

Ferris, D.R. 1995b. Teaching ESL composition students to become independent editors. TESOL Journal. 4(4): 18-22.

Ferris, D.R. 1997. The influence of teacher commentary on student revision. TESOL Quarterly. 31(2): 315-339.

Ferris, D.R. 1999. The case for grammar correction in L2 writing classes: A response to Truscott (1996). Journal of Second Language Writing. 8(1): 1-11.

Ferris, D.R. 2004. The ‘‘Grammar Correction’’ Debate in L2 Writing: Where are we, and where do we go from here?(and what do we do in the meantime …?) Journal of Second Language Writing. 13(1): 49–62

Frantzen, D. 1995. The effects of grammar supplementation on written accuracy in an intermediate Spanish content course. Modern Language Journal. 79 (3): 329-344.

James, C. 1998. Errors in language learning and use: Exploring error analysis. Harlow: Longman.

Johnson, K. 1996. Language teaching & skill learning. Oxford: Blackwell.

Kepner, C. 1991. An experiment in the relationships of types of written feedback to the development of second-language writing skills. Modern Language Journal. 75(3): 305-313

Lalande, J.F. 1982. Reducing composition errors: An experiment. Modern Language Journal. 66 (2): 140-149.

Lee, I. 1997. ESL learners’ performance in error correction in writing: Some implications for teaching. System. 25(4): 465-477.

Littlewood, W. 1984. Foreign and second language learning, language acquisition research and its implications for the classroom. Cambridge: Cambridge University Press.

Loewen, S. 1998. Grammar correction in ESL student writing: How effective is it? Retrieved on June 6, 2003 from http://www.temple.edu/gradmag/fall98/loewen.htm.

Macaro, E. 2001. Learner strategies in second and foreign language classrooms. London: Continuum.

McLaughlin, B. 1987. Theories of second language acquisition. London: Edward Arnold.

O’Malley, J.M. and Chamot, A.U. 1990. Learning strategies in second language acquisition. Cambridge: Cambridge University Press.

Polio, C., Fleck, C. and Ledere, N. 1998. “If I only had more time”: ESL learners changes in linguistic accuracy on essay revisions. Journal of Second Language Writing. 7(1): 43-68.

Robb, S., Ross, T. and Shortreed, I. 1986. Salience of feedback on error and its effect on EFL writing quality. TESOL Quarterly. 20(1): 83-96.

Schmidt, R. 1990. The role of consciousness in second language learning. Applied Linguistics. 11(2): 129- 158.

Semke, H. 1984. Effects of the red pen. Foreign Language Annals. 17(3): 195-202.

Skehan, P. 1989. Individual differences in second language learning. London: Edward Arnold.

Truscott, J. 1996. The case against grammar correction in L2 writing classes. Language Learning. 46(2): 327-369.

Wenden, A. 1987. Incorporating learner training in the classroom. In A. Wenden and J. Rubin. (eds.)

Learner strategies in language learning. Englewood Cliffs: Prentice Hall International, 159-168.

Wenden, A. 1991. Learner strategies for learner autonomy. London: Prentice Hall International.