Is there life after grammar?

Scott Thornbury

This paper was presented at the UCLES CTEFLA Conference, London 1996.

"A language is acquired through practice; it is merely perfected through grammar". Leibniz

Grammar rules

The following comment appeared in a trainee's diary on the first day of a CTEFLA course at our centre in Spain. It was written in response to the question "What are your expectations of this course?"

"At present I anticipate that the course will challenge one's stamina and commitment due to the intensity and concentration on grammar."

While at first this statement suggests a somewhat skewed view of the course, it is not so surprising when you consider the contact the trainee had had with the centre to date. The message conveyed by a lot of the course documentation is that grammar rules. For example, on applying for a place on the course, the first piece of documentation a candidate receives, along with an application form, is a pre-interview task. This consists largely of questions that relate to language systems. Assuming an applicant is accepted for interview, it is these same questions, again, that become an important focus of the interview.

Interviewees who are accepted on to the course are then given a "Pre-Course Task". Again, the bulk of this task is concerned with questions of language (as opposed to either learning or teaching). And when, on Day One of the course, the timetable is handed out, regular grammar (or "Language Analysis") sessions - on such topics as "The Present Simple", "The Futures", and "Modals" - form an important strand of the course.

In addition, some centres, at the outset, provide their trainees with a list of the criteria according to which their success on the course will be judged. One centre I assessed had prioritised the following: "the ability to analyze accurately and present clearly the meaning, form and use of grammar structures". There was no reference, however, to anything along the lines of "the ability to provide opportunities for authentic language use".

This heavy emphasis on grammar, and the teaching of grammar, re-surfaces frequently in course documentation - in the teaching practice points, TP lesson plan rubrics, TP "crit sheet" rubrics, mid-course tutorial forms, etc etc, forming a not-so-hidden agenda. The overriding message that comes through again and again is that the language teacher's main job is not so much to create conditions for authentic language use, but to teach specific linguistic forms.

Grammar teaching vs...

The net effect of this agenda in the teaching practice classes I have observed, both in our own centre and when assessing, is that very often TP students are subjected to lengthy explications of often quite subtle points of grammar mediated through overprolonged sequences of display questions, in which they seldom produce much language beyond the word or phrase level, and rarely engage with the topics or the material. And, at higher levels, where the same old structures have already been presented time and time again, learners are often severely underchallenged.

Tutors, wittingly or unwittingly, are often party to this grammar-driven view of teaching, by focusing, in TP feedback, less on what the students did (and said) than on what the trainees did (and said). The model or metaphor of learning that is being promoted is the "transmission" one: the teacher is the conduit conveying discrete items of knowledge to the (supposedly grateful but totally passive) learners.

This is a view of language learning that, unsurprisingly perhaps, persists beyond the course itself and forms the covert theory of many practising teachers, judging by those candidates who return two or three years later to do the DTEFLA. In fact, it sometimes seems as if the "communicative" revolution had never happened. The whole profession (and publishers must bear some responsibility) seems bent on maintaining the fiction (I will argue it is a fiction) that language teaching involves nothing more than simply teaching the language.

Even academics are implicated, trumpeting, for example, the "return of grammar to the centre stage of language teaching and learning" (Tonkyn 1994: 12). For many, it seems, the recognition that a "focus on form" is a necessary condition for language "restructuring", vindicates "a return to teaching discrete decontextualised grammar points, plus or minus overt grammar explanations", as Larsen-Freeman and Long (1991) put it. They add: "Clearly, we want to avoid an unwarranted inference of that kind". (Larsen-Freeman and Long 1991: 322).

Faint hope! The prevailing winds are grammar driven: language teaching is grammar teaching. And it follows that teacher training is training teachers to teach grammar. (The only thing that seems to have changed is the amount of grammar to be taught: as linguists scramble to describe it, the more there is to teach: not just conditionals 1, 2 and 3, but mixed conditionals; not just words but lexical chunks and collocations; not just written grammar but ellipsis, left-dislocation and tails).

... language learning

But to what extent does a grammar-driven view of teaching accord with what is known about how languages are in fact learned? Very little. As Rutherford (1987) observed: "Theories of grammar ... are not theories of language acquisition". In fact, there is still no conclusive evidence that formal instruction in the grammar is a necessary condition for the learning of a language, even in classrooms. Ellis (1994) sums up the evidence: "There is little, if any, support for the claim that classroom learners must have formal instruction in order to learn the L2". (Ellis 1994: 657; emphasis added). The same scepticism for the value of head-on grammar teaching is taken up in a number of recent publications. Two examples:

We may need to accept that information about the different meanings certain grammatical items can carry may be of more interest and use to grammarians than it is to learners... It is our view that a fundamental implication of all current language acquisition research is that teachers would be wise to remain sceptical of the long-term effects of any kind of formal instruction on the grammatical development of their learners. [Beaumont & Gallaway: 1994: 173]

The nature of our knowledge of language and of language-learning processes is such that...notions of optimally accurate and/or effective rules are neither realistic nor desirable in a learning/teaching situation... It is much more relevant to the learner's needs that a teacher can provide typical language data and monitor learner production effectively (whether for accuracy or for fluency) than the most watertight, static formulations of language data be aimed at. [Westney 1994: 93]

(It's important to underscore the point that none of these writers is suggesting that teachers don't need to "know the rules". Only that they shouldn't try to make these rules the content of their instruction).

So, if grammar teaching doesn't cause language learning, what does? Ellis (1994) goes on to observe:

It is possible, however, that there are certain linguistic properties that cannot be acquired by L2 learners (especially adults), unless they receive instruction in them. ... It is not yet clear which kind of instruction works best but there is evidence to suggest that focusing learners' attention on forms, and the meaning they realize in the context of communicative activities, results in successful language learning. (op. cit. 657-9)

The evidence that he cites comes from researchers like Patsy Lightbown (who co-authored a book called How Languages Are Learned, (OUP 1993) which should be compulsory reading for teacher trainers at all levels). Lightbown found that learning appeared to be optimal in "those situations in which the students knew what they wanted to say and the teacher's interventions made clear to them that there was a particular way to say it" (1991:209), quoted in Ellis 1994: 640). That is, there was feedback "under real operating conditions" (Johnson 1988). Lightbown calls this a "get-it-right-in-the-end" methodology.

Good teachers talk

This suggests an approach that is well captured in statements elementary students made at my centre, when questioned about the qualities of the good teacher:

A good teacher:

  • Does exercises in which the content (which in principle is not the most important part of an English class) is interesting because then the students want to express themselves and speak
  • Makes the students talk and corrects them
  • When the students want to say something and you can't get it out, they help you because when this happens and you end up not being able to say what you wanted you get disheartened
  • Talks to the students

Life after grammar

How might such an approach be promoted at the pre-service level? What follows are some suggestions organised according to the increasing degree of likely change they might entail:

A greater emphasis should be placed on the fourth P - personalisation - within the PPP paradigm. For example, an input session could be devoted to this area alone; and coursebook activities could be critiqued with a view to the opportunities for personalisation that they offer. It should be made clear that personalisation is not an optional extra tagged on to the model, but is the component without which the other components are meaningless - literally.

The enrolment procedures and course documentation should be reviewed to ensure that the agenda conveyed to candidates is not one that prioritises language over learning. At the interview stage, for example, apart from questions about grammar, candidates could be asked about previous learning and teaching experiences: "When did you last teach somebody something? How did you go about it?" Success criteria should be drafted to ensure that "the provision of classroom opportunities for authentic use" is foregrounded.

The notions of "student-centredness" and "communicativeness" should be expanded to mean not simply that students participate and interact. Student-centredness should mean that the lesson content should be, where possible, student initiated and student driven. Communicativeness should be defined qualitatively - e.g. as Puchta and Schratz (1993:3) define it: "If the participants are being both frank and considerate, independent yet cooperative, and are speaking willingly and comprehensibly to particular listeners about things that matter to them both, then the quality of communication is high". The terms STT (student-talking-time) and TTT (teacher-talking-time), traditionally used to measure the degree of "student-centredness" should be scrapped, as they create a false distinction. Teacher talk is as valuable as student talk, if it provides learners with authentic input.

A tendency to want to increase the "language" component of pre-service courses, at the possible expense of "learning" components, should be resisted at all costs. If CLT is to become more communicative it must be supported by an understanding of how language is learned, not described. Rather than the traditional, product-oriented, training syllabus of discrete language items ("Perfect aspect", "Modals", "Narrative tenses" etc), the language component should perhaps be designed in process terms, for example: "Lesson planning: anticipating problems"; "Using reference sources"; "Investigating language transfer" (Kerr 1994). In other words, language awareness should be viewed as "awareness of when linguistic knowledge is what is needed, and the ability to locate, interpret and apply that knowledge. In this sense, linguistics in the content of TEFL training should be regarded less as a content area than as a skill area." (Edge 1988, emphasis added).

The programme could be re-designed so as to position fluency-type activities (or authentic use activities [Scrivener 1995]) sooner in the course, and to postpone the introduction of presentation techniques, on the assumption that (a) this may serve to attach greater importance to authentic use; and (b) presentation techniques, involving as they do a minimal degree of language analysis skill, are relatively late acquired.

Offering an alternative to the (grammar-driven) PPP model and developing the requisite management skills that would realise it. Consistent with Lightbown's finding, the essential skills for a "get-it-right-in-the-end" methodology are a) setting tasks that produce learner output; and b) providing feedback on this output. A brief taxonomy of such skills might include the following:

  • Task-setting skills for output
  • Asking questions; "real" questions vs display questions; nominated vs open questions
  • Eliciting: words, phrases, full sentences
  • Student-initiated questions
  • Setting up teacher->student and student->teacher exchanges
  • Setting up open and closed pairs work
  • Dialogue building
  • Setting up group work and report-back
  • Setting up survey/questionnaire (milling) type activities + report
  • Setting short communicative writing tasks
  • Setting up extended speaking activities, e.g. "show and tell", simulations and role plays

Feedback skills

  • feedback on content
  • feedback on form: negative feedback
  • clarification requests
  • rephrasing/reformulating
  • error analysis
  • instruction and explanation (including rule formulation)

These, then, are the skills that could be foregrounded in initial teacher training. (I am aware that many courses do in fact emphasise the importance of these skills - but often only as "practice" techniques within the presentation-practice paradigm).

Note that feedback skills 1 - 4 make no requirement of analysis; feedback skills 2 - 4 require, however, that the teacher can recognise what is "correct" or not, or, more specifically, what is a) intelligible; b) accurate; c) appropriate. I suggest that some of the time traditionally taken up on initial training courses with "language analysis" could usefully be spent on "error analysis" instead, including "grammaticality judgement tests": for example, playing trainees recordings of learner output and asking them to indicate when they hear an error. This could be followed up with more detailed analysis of transcripts of the recordings.

Notice, also, that "instruction and explanation" is a "late-acquired skill", in that it assumes the ability not only to analyse learners' mistakes but to provide corrective "rules" - or, more usefully, "rules of thumb" (see Swan 1994). At this stage learners could be introduced to "presentation" techniques that - in more traditional courses - are dealt with much earlier in the programme.

 

References

Beaumont, M. and Gallaway, C. 1994 'Articles of faith: The acquisition, learning and teaching of a and the' in Bygate et al (eds.) 160-74.

Bygate, M., A. Tonkyn, and E. Williams, 1994 Grammar and The Language Teacher. Hemel Hempstead: Prentice Hall.

Edge, J. 1988 'Applying linguistics in English language teacher training for speakers of other languages' in English Language Teaching Journal 42:1, 9-13.

Ellis, R. 1994 The Study of Second Language Acquisition. Oxford: Oxford University Press.

Johnson, K. 1988 'Mistake correction' in English Language Teaching Journal 42:2, 89-96.

Kerr, P. 1994 'Language awareness in native-speaker teacher training courses' in The Newsletter of the Special Interest Group for Teacher Trainers of IATEFL No. 11, 3-5.

Larsen-Freeman, D., & Long, M. 1991 An Introduction to Second Language Acquisition Research. Harlow: Longman

Odlin, T. (ed.) 1994 Perspectives on Pedagogical Grammar. Cambridge: Cambridge University Press.

Puchta, H. and M. Schratz 1993 Teaching Teenagers Harlow: Longman.

Rutherford, W.E. 1987 Second Language Grammar: Learning and Teaching. Harlow: Longman.

Scrivener, J. 1995 Learning Teaching Oxford: Heinemann ELT

Swan, M. 1994 'Design criteria for pedagogic language rules' in Bygate et al.

Tonkyn, A. 1994 'Introduction: grammar and the language teacher' in Bygate et al. (eds.) 1-14

Westney, P. 1994 'Rules and pedagogical grammar', in Odlin (ed.) 72-96

 

More articles on language teaching