A Lifetime of Grammar Teaching

Rod Ellis

University of Auckland

(This article was shared in the Instructed Second Language Acquisition Class at Anaheim University, 2013).

Linh Phung's Dissertation Defense

I have worked as a language teacher, teacher educator and second language acquisition (SLA) researcher for over forty years. During this time grammar has figured large in my thinking, in part because it has traditionally been so central to language pedagogy and in part because I became fascinated with how the human mind grapples with the task of learning the grammar of a second language (L2).  Throughout this period I have sought to address what I still see as the key question – How can the grammar of a second language be taught in a way that will maximize the chances of students learning it?

My life as a teacher

I began my career as a ‘backpacker English teacher’ in a Berlitz school in Spain where I received a few days training in the ‘Berlitz Method’ (a version of the then popular direct method). This method seeks to elicit correct grammatical sentences from students inductively (i.e. the teacher is expressly forbidden from providing metalinguistic explanations of the target structures). It aims to achieve functional ability in English by ensuring that students developed correct grammatical ‘habits’. I recall the frustration that I and my students felt when, despite my best attempts to help them produce the target structure correctly, they frequently failed to do so. After four months, I had had enough of my Berlitz experience and returned to England. However, this period was formative. I found that although I did not enjoy the Berlitz Method I did enjoy teaching. 

There was an interim year before I took off on my travels again when I worked as a primary school teacher in a school in London. Then in January 1967 I left to work in a newly established secondary school in Zambia. Zambia had only very recently achieved independence from Britain and was prioritizing secondary school education as a central element in nation-building. The English curriculum that I and the other teachers developed was built around the two central notions about learning English - ‘skill-getting’ and ‘skill-using’ (Rivers and Temperley, 1978). Although, the term ‘communicative language teaching’ had not yet appeared on the scene, the curriculum we devised and taught in fact conformed to many of the principles of this approach. ‘Skill-getting’ involved helping students develop the ‘knowledge’ of English they needed to communicate effectively and involved the intensive practice of the sounds, vocabulary and grammar of English. ‘Skill-using’ involved providing opportunities for the students to use their knowledge of English to communicate orally and in writing. Our curriculum consisted of a careful balance of activities that catered to these two aspects of learning.

Grammar was central to the ‘skill-getting’ component. We worked from a structural syllabus that listed the grammatical structures to be taught in a structure-a-day approach. Although I was not aware at the time, this was the period during which the ‘language teaching controversy’ (Diller, 1978) was prominent. There were those who espoused an empiricist, structural view of language (e.g. Bloomfield and Fries) and those who viewed language in rationalist terms (e.g. Chomsky). In language teaching, there were advocates of the audiolingual method (a development of the direct method I was already familiar with from my Berlitz experience) with its emphasis on mimicry, memorization and pattern drill and other advocates of the cognitive code method, which emphasized the psychological reality of ‘rules’ and the importance of presenting these clearly and explicitly to learners.

We experienced problems with both approaches. Here is an anecdote that illustrates the kind of problem we experienced with the audiolingual approach. I was attempting to address errors involving verbs of possession (e.g. * I am having two brothers).

After extensive oral drilling during which the students successfully produced correct sentences, I set the students a written exercise and then noticed that there was one student at the back doing nothing. When I approached him and asked him why, he replied ‘I am not having my exercise book’. Clearly, there had been no transfer of learning from the drill to communicative use! The cognitive code method raised a different kind of problem. While some of the Zambian students benefited from the systematic and detailed explicit explanations of grammar points, many did not. The emphasis on grammatical form and an analytical approach to language seemed to run counter to their natural inclination to treat language as a communicative tool. Thus, neither audiolingual drills nor cognitive explanations of grammar proved very successful.

Together with Brian Tomlinson I began to try out a different approach known as the ‘oral-situational method’. This espoused the importance of studying language in relation to the ‘context of situation’. That is, grammar was treated as a resource for realizing meanings in relation to the specific situations. Applied to language teaching, this meant that grammar should not be taught as a set of formal patterns or abstract grammar rules, but rather through situations that would demonstrate the meanings realized by specific grammatical forms. The oral-situational method differed from audiolingualism in that it viewed grammar as a means for realizing meanings rather than as a set of habits while it differed from the cognitive-code method in that it placed much less emphasis on the explicit explanation of grammar points.

As there were no textbooks based on the oral-situational approach suitable for the Zambian context, Brian and myself set about constructing our own grammar teaching materials. These were eventually published in 1973 as English Through Situations by Longman. The materials attempted to practice specific grammatical points in relation to situational information that demonstrated their meanings. They also sought to make grammar more interesting for the students by including activities that were humorous and, sometimes, zany. I recall the publisher censoring some of our more risqué ideas. For example, we invented a character called ‘Mr Cabbage’ but, reluctantly, agreed to change this name to the more mundane ‘Mr Mwenda’. 

The materials we developed reflected a number of general principles about the teaching of grammar. One was that teaching grammar should involve input-based as well as production-based activities. Another principle – central to the oral-situation approach – was that grammar consists of form-meaning mappings and that ‘correctness’ involves the use of a specific form to express a specific meaning.  A third principle is the ‘real operating conditions principle’ according to which learners need to process grammatical forms in the same conditions in which they will experience them in real communication.  However, I was not consciously aware of these principles at the time we were writing English Through Situations. They emerged in explicit form much later as a product of my work in SLA. At the time, my thinking about grammar was based on a view of language (i.e. what grammar is) rather than on any notion of how grammar is actually learned.

There was, of course, much more to the curriculum we taught than grammar. Many lessons were directed at ‘skill-development’. We developed a classroom library system, taking boxes of carefully selected graded readers into our classrooms twice a week and encouraging the students to read at least one book per week. This was long before Stephen Krashen was to promote extensive reading as a primary source of ‘acquisition’. We also had class readers (i.e. each student had a copy of the same book) which we read aloud to the students and then used as a source for various fluency activities. Recognizing that many of our students were extremely slow readers (some were reading no faster than 30 words per minute) we instituted faster reading lessons to train students in the techniques of rapid reading.  We conducted intensive reading lessons using John Munby’s (1968) Read and Think, which used carefully designed multiple choice questions to elicit small group and class discussion of the different levels of meaning in a text. Much of what we did would qualify as what is now called ‘task-based language teaching’. We saw such work as an important complement to the ‘skill-getting’ lessons where grammar was taught.

The three years I spent at Kaoma Secondary School provided me with a truly concrete experience of language teaching. They whetted my appetite to learn more about language, language use and language learning.  In 1970 I returned to England to study for an MA at the University of Leeds.  

Postgraduate courses in TESOL or applied linguistics are now very common but in 1970 they were still a relatively new phenomenon. The Leeds programme offered courses in transformational generative grammar, field linguistics, sociolinguistics, psycholinguistics and language teaching. This was a time of paradigm change in linguistics, with the structural school of Bloomfield and Fries giving way to the generative school of Chomsky and with the corresponding shift from behaviourism to nativism in the psychological underpinnings of linguistic theory. It was an exciting time. However, the application of these new ideas to language learning and teaching was still in its infancy and was only weakly reflected in the content of the Leeds programme. My chief memory is of a lecture on Newmark’s  (1966) ideas about language teaching. Newmark  argued that we should not ‘interfere’ with language learning as this will take place naturally without the need for teachers to present and practice specific linguistic items. I found this proposal interesting but it was not till sometime later when I began to learn more about L2 acquisition that I took it seriously.

My life as a teacher educator

After my MA I returned to Zambia to take up a lecturer position in Nkrumah Teachers’ College in Zambia. Altogether I spent five and half years at the College - years that in many ways were as formative as the years at Kaoma Secondary School. They were especially memorable because I was able to link up again with Brian Tomlinson.  The lectures on various aspects of teaching that we gave to the teacher trainees eventually formed the basis for a book that we jointly authored, Teaching Secondary English: A Guide to Teaching English as a Second Language, which was eventually published by Longman in 1980. Three of the chapters in this book addressed grammar in one way or another and I will focus on these as illustrative of my thinking at this time.

The chapter entitled ‘Structure’ consisted of descriptions and examples of different types of grammar activities. These were organized according to the approach we had adopted in Teaching English Through Situations. That is, they started with ‘mechanical exercises’ and then moved on to ‘situational exercises’. There was also a section on planning structure lessons, where the traditional present-practice-produce (PPP) format was recommended, and another on actually teaching a structure lesson, which dealt with such aspects as how to conduct oral practice and deal with errors. The chapter concluded with some suggestions for ‘follow-up lessons’ (in recognition that even the most perfectly executed lesson might not succeed in ensuring correct use of a target feature). By and large, these chapters reflected received opinion regarding how to teach grammar in the late 1970s and are still evident in mainstream opinion about grammar teaching today.

Looking back at Teaching Secondary English, I can see the views of about language learning that underpinned our recommendations for teaching grammar. Interestingly, the chapter on ‘Structure’ began by distinguishing two types of knowledge of grammar – ‘unconscious knowledge’ and ‘acquired knowledge’. The former referred to ‘the unconscious ability to use the structure of a language to convey meaning’ and the latter to ‘information that has been deliberately acquired through studying structural descriptions’ (p. 8). We emphasized the importance of ‘unconscious knowledge’ on the grounds that this was needed ‘to produce sentences automatically’. However, we also argued that ‘it is often useful to refer the pupils to simple grammatical descriptions as a learning aid, i.e. as a means of developing ‘conscious knowledge’. Later, I realised that this distinction mirrored Krashen’s (1981) distinction between ‘acquired’ and ‘learned’ knowledge.

Another important distinction to be found in the ‘Structure’ chapter is that between ‘initial teaching’ and ‘remedial teaching’. The former assumes that the pupil has no knowledge of the structure to be taught, while the latter aims to address errors that learners continue to make in a structure that has been previously taught. We attached considerable importance to remedial teaching on the grounds that secondary school students would have already been taught most structures during their primary school years but may not have mastered them. In fact, we included a separate chapter on remedial teaching. In this we recommended that teachers should not blindly follow the structural syllabus but rather first endeavour to find out which items on this syllabus their students were already able to use correctly so that they could focus on those that were continuing to manifest errors. We gave detailed suggestions for how to conduct remedial lessons (e.g. developing a bank of correction cards which could be given to individual students when they committed specific errors). Implicit in our views about remedial teaching was the belief that learning the grammar of an L2 is a slow and gradual process that will be characterized by errors which will only be eliminated over time. Explicit was our view that grammar teaching could assist this process and made some attempt to address this thorny issue in the sections of Teaching Secondary English that addressed how teachers should respond to learner errors. We advanced the view emanating from early work on error analysis (e.g. Richards, 1974) that ‘it was desirable for ESL learners to make errors’ because these served as the means by which they tested out hypotheses about the target language. However, we made no real attempt to show how the teaching of grammar could be reconciled with such a view of L2 learning. Our solution to the problem of recurring errors was simply ‘more teaching’.

Also of interest is how we handled error correction. In the case of oral drills, we recommended that the teacher should clearly signal that an error had been committed and then repeat the stimulus, thus pushing the student to self-correct. We also pointed out that if students persisted in making a particular error, it was necessary to stop the drill and provide a brief explicit explanation of the target feature. We were adamant that ‘allowing errors to pass uncorrected negates the whole purpose of drilling’ (p. 39). These views reflected those that were widely expressed in other teacher handbooks at that time (and indeed still are today). They did not draw on any empirical evidence of how teachers actually carry out corrective feedback or on the effect that such practices have on L2 acquisition.

In writing Teaching Secondary English Brian and myself drew both on our prior experiences of teaching in secondary schools in Zambia and also on the ‘technical knowledge’ we had obtained from our masters courses and our reading of applied linguistics texts such as Richards (1974). Looking back, there seems one glaring omission in how we conceptualized grammar teaching already hinted at in my comments about how we viewed error correction. We saw language teaching as a set of externally defined practices (drills, situational exercises, correction cards etc.) rather than as interactions that teachers and students engage in. I think we had very little sense of teaching as a classroom process and little idea of how this process created the contexts through which learning could take place.

My life as a researcher

When I returned from Zambia in 1978 I immediately started another masters course – this time an Master in Education at Bristol University. I studied with Gordon Wells and his co-researchers, who were involved in a longitudinal study of child first language acquisition – the Bristol Study of Language Development. The project drew heavily on earlier research on child language acquisition (e.g. the work undertaken by Roger Brown in the United States) but extended this in two major ways. First, in addition to examining the children’s development of grammar, it also addressed the functional properties of their use of language, drawing on Michael Halliday’s functional grammar. Second, it investigated the relationship between different properties of caretakers’ talk and the children’s language development. It was this latter aspect of the project that I worked on producing a thesis entitled ‘Enabling Factors in Caretaker-Child Discourse’. The Bristol course introduced me to research on child language acquisition and gave me the opportunity to cut my teeth as a researcher - to work with empirical data, to devise systems for analyzing it, and to learn how to use statistical tools. It changed my professional life. From that time onwards I wanted to be a researcher.

On completing my MEd, I took a position at St Mary’s College in London. More or less at the same time I enrolled to study for a PhD in the Institute of Education, University of London. The three years I spent studying for my PhD were, in many ways, the most stimulating and exciting of my professional life. The key idea that informed my doctoral thesis was that L2 acquisition arose out of the discourse that learners participated in. My starting point was to familiarize myself with the research that had investigated L2 acquisition. In 1979 this was still quite limited as ‘SLA’ at that time was still a very young field. I was struck by two major findings of the early research. First, like L1 children, L2 learners learned the grammar of the language gradually and incrementally, manifesting orders and sequences of acquisition that were remarkably uniform irrespective of their age or their first language. L2 learners had their own built-in syllabus, which Selinker (1972) called ‘interlanguage’. Second, like L1 children, L2 learners (both children and adults) participated in conversations with other speakers of the language, which afforded them the data they needed to activate their syllabus. Hatch (1978), for example, proposed that :

One learns how to do conversations, one learns how to interact verbally, and out of this interaction syntactic structures are developed.

In short, I found clear support for the idea that motivated my thesis in the SLA literature. However, the bulk of the research had investigated naturalistic learners who were living in a country where the L2 was spoken and who therefore learned the L2 through the everyday conversations that were part of their experience of living. This led me to ask whether L2 classroom learners manifested the same order and sequence as naturalistic learners and in what ways the interactions they experienced in the classroom facilitated (or impeded) their L2 development. The research I undertook for my thesis involved a two-year longitudinal study of three ESL learners who were more or less complete beginners when I began the research.  I collected data from within the classroom. My analyses examined the learners’ use of formulaic speech, the order and sequence of development of negatives and questions and the nature of the different classroom interactions they participated in. At one stage, I conducted an experimental study in which the class the three learners were in received instruction in the formation of WH-questions, which my data showed the learners had not yet acquired. The thesis led to a number of publications, including a book published by Pergamon in 1984 with the title Classroom Second Language Development. In it, I advanced the idea that classrooms had to be seen as places where grammar was not ‘taught’ but rather ‘acquired’ through the interactions that took place there.

In 1984 I took up the position of Head of the ELT Department in Ealing College of Education (later to become Thames valley University). I continued to write – Understanding Second Language Acquisition was published by Oxford University Press in 1985. I also continued to research L2 acquisition and to write a number of papers. Drawing on data from my doctoral thesis, I focused on two key issues – the role of interaction in L2 learning and the variability of learner language – both of which were explored in relation to how learners acquired the grammar of an L2.

In my work on interaction, I sought to show how the conversations in which learners participate in the classroom provide them with units of different sizes which they then incorporate into their own utterances and how this can explain the incremental nature of L2 acquisition. I built on my view that language teaching should be viewed not so much as a set of techniques for imparting knowledge about the language but as ‘interaction’.  It followed that, in this sense, the learning that goes on in an instructed environment may not be so different from the learning that takes place ‘naturalistically’. My work on variability was motivated by the fact that L2 learners manifest enormous variation in their use of an L2 depending on such factors as whom they are addressing, the linguistic context of the grammatical structures they are attempting to use, and the extent to which they are able to plan what they want to say before executing an utterance. This work reinforced my view that errors were natural and inevitable and, furthermore, that they were largely systematic, reflecting the way in which learners gradually constructed their interlanguages.

While at Thames Valley University I also embarked on a new research project entitled ‘The Acquisition of German in a Classroom Context’. This project investigated the extent to which classroom learners manifested the same acquisition sequences in grammar as naturalistic learners.  It found that, by and large, they did but that they seemed to advance more rapidly along these sequences than naturalistic learners.  The project was also the start of my ongoing interest in individual learner differences. It was my first attempt to undertake qualitative research as I asked five learners to keep a diary of their learning experiences over a six month period. My experience with this project convinced me of the merits of research that drew on both quantitative and qualitative traditions.

In 1988 I was invited to conduct a seminar on second language acquisition for Temple University Japan (TUJ). This proved another landmark in my professional career as it led to me eventually taking up a four year position as professor at TUJ where, for the first time, I had the opportunity to supervise doctoral students. A number of the students I supervised produced exemplary dissertations. It is not possible to mention all of them but I would like to briefly summarise the research of three. Tomoko Kaneko examined the use of English and Japanese (the students’ L1) in junior college English lessons. She was also interested in what the students were able to report having learned from these lessons and the types of interactions that promoted this. One of her main findings was that the teachers in the classes she investigated used far more Japanese than English! The second student to finish was Sandra Fotos. She drew on my (and others’) ideas about consciousness-raising to investigate the extent to which this was effective in helping learners learn grammatical structures in English. She was able to show that consciousness-raising via grammar discovery tasks worked as effectively as direct grammar explanations and also that the knowledge that learners gained from these tasks facilitated their subsequent noticing of the grammatical structures in input to which they were exposed. Gordon Robson examined the relationships among students’ participation in English lessons, their language proficiency and individual difference factors (personality and language anxiety). He found that extraverted learners and more orally proficient learners participated more and anxious learners less.

At TUJ I was able to use my understanding of SLA to put together my ideas about grammar teaching. These ideas were based on one central assumption: for teaching to work it must accord with how learners learn grammar. This assumption originated in my experiences of teaching grammar at Kaoma Secondary School, but my ideas about how to implement it only gradually solidified as a result of my work in SLA. Perhaps the key article I wrote during this period was ‘Second language acquisition and the structural syllabus’, which appeared in TESOL Quarterly in 1993.  In this, I tried to grapple with what many SLA researchers, myself included, saw as the central issue in grammar teaching – how to organize a grammar syllabus in a way that took account of the learner’s own built in syllabus. I argued that a structural syllabus could not serve as a basis for developing learners’ implicit knowledge (i.e. the kind of intuitive knowledge needed for easy communication) because the acquisition of this type of knowledge was not a matter of ‘accumulated entities’ (Rutherford, 1987) but was an organic, learner-driven process. I also noted the difficulty of trying to match the teaching syllabus with the learner’s syllabus given that teachers could not be expected to know precisely where each student was at. I concluded that, where implicit knowledge was concerned, a structural syllabus was simply not viable. However, such a syllabus could be used as a basis for developing learners’ explicit knowledge (i.e. declarative knowledge of grammar ‘facts’).  I then went on to argue that explicit knowledge plays a facilitating role in grammar learning by enabling learners to ‘notice’ exemplars of grammatical features in the input, to ‘notice-the-gap’ between what they noticed and their own erroneous production of these features, and to monitor their output by self-correcting their errors.

In short, the theory of grammar teaching I developed proposed that instruction should be directed at promoting the cognitive processes that were involved in the development of implicit knowledge but should not attempt to teach implicit knowledge directly as the acquisition of this type of knowledge was controlled by the learner. By and large, this is a theory to which I still subscribe. It has not been without its critics, however!

This theory was directed at explaining how instruction could work for ‘new’ grammatical features (i.e. features that had not yet begun to enter the learner’s interlanguage). However, not all grammar learning involves the acquisition of ‘new’ features. It also involves the development of control over features that have already been partially acquired (i.e. features that are being used by the learner variably). To address this, I drew on Merrill Swain’s ideas about ‘comprehensible output’. This staked out a role for learner production by proposing that learners did not just learn from input but also by being pushed to use the L2 more concisely and more accurately. In a small-scale study with Junko Nobuyoshi, learners received requests for clarification when they made a past tense error in their oral narratives. We investigated whether this resulted in the students correcting their own errors. We were also interested in whether this treatment enabled them to use the past tense more accurately in a subsequent oral narrative when they were not ‘pushed’ to correct themselves. We found that for some of the learners did correct their errors and there was also subsequent improvement in accuracy. It showed that, under certain conditions, output-based grammar instruction could be effective.

During my time at TUJ, I continued to write. The Study of Second Language Acquisition (Oxford University Press) was published in 1994. I worked with Sarah Murray on writing a textbook (Lets Use English) for primary school learners of English in South Africa, visiting that country several times in the process. Textbook writing has continued to figure in my professional work throughout my career. I have always seen it as an important element of an applied linguists’ work as it serves as a means of creating the link between theory and the practice of language teaching. Lets use English drew on many of the ideas about language teaching that I had developed through my study of SLA.

In 1993, I moved from TUJ to the main campus of Temple University in Philadelphia. Philadelphia itself did not prove to be one of the geographical highlights of my peripatetic career but I hugely enjoyed the work in the University and the students there. I continued to write academic books - Second Language Acquisition and a collection of my previously published papers in SLA Research and Language Teaching, both published by Oxford University Press in 1997. Grammar learning and grammar teaching figured largely in both these books and also in an article called ‘Teaching and research: Options in grammar teaching’ published in TESOL Quarterly in 1998. In this article I discussed a number of options for teaching grammar – input-based instruction, explicit instruction, production-based instruction and corrective feedback – and tried to show how each of these could be informed by research in SLA.

I also worked on a grammar textbook called Impact Grammar (Pearson Longman, 1999) with Stephen Gaies. This is the textbook that best illustrates my attempts to implement my theoretical ideas about grammar teaching. The book was premised on the idea that being ‘aware’ of a grammatical feature (either in the sense of ‘noticing’ or ‘understanding’) is the not the same as ‘acquiring’ the feature but that awareness can facilitate grammar acquisition. The book also aimed to assist students to become autonomous learners of grammar through the use of discovery activities. Each unit dealt with a different grammatical point which SLA research had shown was likely to be problematic to L2 learners. There were activities in each unit: Listening for Meaning (where students listened to a short text containing the target structure just to understand the meaning), Listening to Notice (where they listened to the same text to attend to the exemplars of the target structure, Understanding (where they were guided into the formation of an explicit rule for the structure), Checking (where they applied the rule to some new data) and Trying It (where they completed a short task using the structure). The book was intended as a remedial grammar course that could be used either for self-study or in a grammar course.

In 1998 I joined the University of Auckland as Professor in the Institute of Language Teaching and Learning and helped set up a new department – the Department of Applied language Studies and Linguistics. In conjunction with other members of the department I completed four research projects all related in one way or another to grammar learning and teaching. The first investigated ‘focus on form’, a term used by Long (1991) to refer to the incidental attention paid to linguistic form while learners are engaged in performing communicative tasks. We identified the ‘form-focused episodes’ (i.e. episodes where the teacher and the students departed from the main task to address a problem with a linguistic form that had arisen) in transcripts of task-based lessons. We discovered that both the learners and the teachers often paid implicit and explicit attention to form even though the primary focus was on communicating.

The second research project was funded by the New Zealand Ministry of Education. I was invited to undertake a review of the literature on instructed L2 learning. This was eventually published and led to an article – ‘Principles of instructed language learning’ published in System in 2003.  There was also a follow-up to this project – ‘Instructed Second language Acquisition: Case Studies’. This project was carried out together with Rosemary Erlam and Keiko Sakui. It made use of the principles I had identified in my review as a basis for carrying out case studies of four experienced teachers of foreign languages in New Zealand secondary schools. We devised a classroom observation scheme and a set of interview questions with a view to describing both the practices and belief-systems of the four teachers. The idea was to provide an evidence base in the form of illustrative material and exemplars of effective practice that could be used to assist teachers’ professional learning and practice. It was an attempt to help teachers connect theory and practice.

The third project was an ambitious one. Cathie Elder and I received a grant from the Marsden Fund, a contestable fund administered by the Royal Society of New Zealand. The project aimed to develop tests of implicit and explicit L2 knowledge, to examine the relationship between measures obtained from these tests and measures of general L2 proficiency, and to use the tests to investigate the effects of different kinds of grammar instruction. Focusing on grammar, we developed and validated a battery of tests designed to provide relatively separate measures of the two knowledge types by administering the tests to samples of both native speakers and L2 learners. We explored the relationship between these measures and the learners’ IELTS and TOEFL scores. Finally we used the tests to investigate the effects of different types of grammar instruction on learners’ acquisition of implicit and explicit knowledge. The project resulted in a full-length book – Implicit and Explicit Knowledge in Second Language Learning, Teaching and Testing – published in 2009 by Multilingual Matters.

The fourth project investigated written corrective feedback. Together with Younghee Sheen, Hide Takashima and Mihoko Murakami I investigated the effects of focused versus unfocused feedback on Japanese university students’ use of English articles in writing. Sheen carried out a parallel study on adult ESL learners in the United States. The resulting articles (both published by System) have contributed to the lively and ongoing debate that has taken place in the field of L2 writing studies regarding the role that written corrective feedback plays in L2 acquisition. Both studies suggested that written corrective feedback did have an effect on improving the accuracy with which learners used grammatical structures in their writing.

Looking back on my 12 years at the University of Auckland, I can see that in many respects they have been the pinnacle of my professional career. They have afforded me the opportunity to contribute to the development of a new department, to supervise some extremely able doctoral students, to undertake a number of research projects, and to continue to write.

Conclusion

My professional career has followed the traditional path of many British and some North American applied linguists. I began as a teacher of English as a foreign and then second language. I moved on to working as a teacher educator. Then, out of these experiences, I developed skills as a researcher and as a writer.

How grammar is taught and learned have been key themes throughout my professional life. Many of the articles I have written over the years reflect this concern. I will conclude with some general statements about what I have learned about grammar teaching.

1. Learners do not always learn what teachers teach them, especially if by ‘learn’ we mean the kind of implicit knowledge need when communicating.

2. Learners can learn grammar without any explicit grammar teaching. They learn it best by participating in interactions that make the use of specific grammatical features functionally important for communicating.

3. To learn grammar when communicating, learners need to ‘notice’ grammatical features in the input and be ‘pushed’ to use those they have partially learned correctly.

4. Learners benefit from the corrective feedback they receive when they make errors while communicating.  This feedback can be both of the implicit and explicit kinds.

5. Explicit grammar teaching is most useful when it is directed at raising learners’ awareness about grammatical features. This awareness can subsequently help them with the processes involved in developing implicit knowledge.

6. Grammar teaching needs to be broadly conceptualized as involving different of methodological options that include not just the presentation and practice of a grammatical feature but also input-based approaches and task-based approaches.

My professional life has taught me how research can help to modify and clarify the ideas about grammar teaching that were forged initially through the practical experience of teaching.

References

Diller, K. (1978). The language teaching controversy. Rowley, Mass.: Newbury House.

Hatch, E. (1978). Discourse analysis and second language acquisition. In E. Hatch (Ed.), Second language acquisition. Rowley, Mass.: Newbury House.

Krashen, S. (1981). Second language acquisition and second language learning. Oxford: Pergamon.

Long, M. (1991).  Focus on form: A design feature in language teaching methodology. In K. de Bot, R. Ginsberg & C. Kramsch (Eds.), Foreign language research in cross-cultural perspective (pp. 39–52). Amsterdam: John Benjamin.

Munby, J. (1968). Read and think: Training in intensive reading skills. London: Longman.

Newmark, L. (1966). How not to interfere in language learning. International Journal of American Linguistics, 32, 77–87.

Richards, J. (Ed.). (1974). Error analysis. London: Longman.

Rivers, W. & Temperley, M. (1978).  A practical guide to the teaching of English. New York: Oxford University Press.

Rutherford, W. (1987). Second language grammar: Learning and teaching. London: Longman.

Selinker, L. (1972). Interlanguage. International Review of Applied Linguistics, 10, 209–31.