How often have you heard teachers in your English department complain out loud at meetings that many students in their classes, "still can't speak English very well"? How often have you heard the same teachers continue, "But we're investigating this issue and we'll bring you our results and conclusions soon."? The answer to my first question is probably "any day of the school year". The answer to my second question is, I am equally sure, "never".

So the problem of alleged oral inadequacy continues. Another case of complaints without action which, I’m sure you’ll agree, is a very sorry state of affairs. This is generally because language teachers feel tightly-bound by their traditional classroom role as "pedagogues", - "providers of information on formal rules". They do not generally see themselves as "reflective practitioners" who might benefit from observing other teachers and, in turn, being observed.

The term "reflective practitioner", first used by Schon (1983), describes teachers who seek to:

"...discover more about their own teaching by seeking to understand the processes of teaching and learning in their own and others' classrooms", quoted in Wajnryb (1992:9).

In this article I would like to outline:

• how Teacher-Development sessions, when extended to exploratory practice, provided insights into what puzzled us about our Intermediate level students' oral ability.

• how, due to our insights, chosen pair-work topics and tasks helped overcome the limitations on oral work brought about by an impoverished, traditional classroom environment.

The "we" referred to in this article involved a native and a non-native speaker, each with several years' experience with Intermediate level students, plus one non-native speaker in her initial teaching year. The three of us had very busy timetables but shared three Intermediate level classes: 7A, 7B and 7C. In our weekly (40 minute) Teacher Development session we agreed that our professional development would benefit by peer-observation and sharing feedback on the lesson. In this way Teacher Development could be extended to Exploratory Practice, which Allwright and Bailey (1991:2) describe as, "a way in which teachers can investigate their classrooms by gathering information and interpreting it in order to better understand what occurs".

We felt that peer-observation would help us to:

• sharpen our understanding of what puzzled us

• select an issue to work on.

• find appropriate classroom procedures to assist our exploration.

Teachers new to the area of Exploratory Teaching, as we were, will find a vast amount of introductory help in 'The British Council Newsletter Supplement', containing a "Puzzle Page", by Ann Haznedar and Denise Özdeniz, (February, 1997).

The most puzzling issue facing us with 7A, 7B and 7C was the complaint referred to in the introduction to this article. Stated more explicitly, "Our students still can't speak English very well", became:

"Considering the amount of language lessons the three classes had received, even the most highly motivated students seemed incapable of producing, or unwilling to do so, very much orally beyond brief but accurate statements about the content of the current lesson".

I'm sure that we are not alone in articulating such a complaint. Other schools probably face similar problems. We felt, though, by reflecting on this issue, we were displaying healthy professional dissatisfaction rather than petty cynicism, and believed exploratory practice would help.

Our response to the puzzle, over one school term, was three-fold. It first involved sitting and observing, at the back of each others’ classroom. The next stage was to make audio-recordings and analyze transcripts - written records of what was said - of these lessons.

All three tasks are not without problems and teachers who intend to engage in exploratory teaching might find it useful to read my comments below.

One of the most important obstacles to peer-observation is negotiating access; not everyone likes to be watched whilst teaching ! In our case this wasn't such a sensitive area since we were acting as a professional, mutual-support group.

Our initial attempts at exploratory practice (as was the case with the early history of classroom research) involved ticking boxes. When a student or the teacher uttered a relevant remark, meant to be heard by all in the class, a tick was placed next to his/her name. The length of the remark was indicated by one, two, or three ticks.

This "quick and dirty" analysis, as it is commonly called, was a useful starting point. It highlighted the main drawback to teacher-fronted or "lock-step" lessons: in almost every lesson nearly all of what was said stemmed from the teacher. A glance at our check-list showed that teacher-talking time (TTT) was usually above 80%. In our classes of 25 students, on average, 17 students said nothing at all in the target language during a 40 minute lesson.

The next stage of exploratory practice was to make our problem even more explicit. We wished to see how much TTT was occurring compared with student talking time (STT). We also wished to discover just what teachers and students said to each other in the classroom. However, attempts to record lessons involve what Anthropologists call the "observers' paradox". It is also known as ‘reactivity’ and means an alteration in the normal behaviour of a subject under observation, due to the observation itself.

• Should we explain to the class why the lesson was being recorded and risk having the data distorted by students becoming over-talkative in the presence of recording equipment ?
• Should we record in secret and, in doing so, invade students’ privacy and abuse their trust in us?

Allwright and Bailey (199:68) state: "..this major headache can be overcome with patience. Repeated visits to the classroom will familiarize students and teacher with taping and equipment".

Due to time constraints we opted for the middle ground by taping in secret, using a simple cassette-recorder with an in-built microphone. After playing the recording for the group to hear, we asked their permission to use it as data. Younger students are usually keen to hear themselves and to be heard by friends. Future researchers should be warned though: other groups of learners might not be so obliging! (For further comments on such "deception", and many other matters on classroom research see Allwright and Bailey, 'Focus on the Language Classroom', (1991:70/71) which is a key text for teachers looking for more detailed, but extremely readable, insights into what classroom research is all about).

Listening to audio-recordings of lessons, although interesting at first, does not reveal very much. The next stage involved writing transcripts of the lessons we had taped. Transcribing even one forty minute lesson in detail can be a phenomenal task. It involves repeated listening and annotating but the end-result highlights patterns of interaction that check-lists cannot locate. Allwright and Bailey (1991:Appendix H) helpfully list the most useful transcription conventions. They also state that the law of least effort should be applied to transcribing lessons: use only the conventions needed, otherwise a one hour tape can take up to twenty hours to transcribe if done absolutely accurately, including gaps and overlaps.

When the person transcribing has not actually taught the lesson involved, it is useful to confer with him / her after the transcription process is complete. This enables "inter-observer agreement" to take place. By also involving learners, even more reliable "triangulation of data" would occur. This term, borrowed from land-surveying, suggests that at least two perspectives are necessary if an accurate picture of a particular phenomenon is to be obtained.

The transcripts of lessons with 7A, 7B and 7C reflected the general aspects of teacher-fronted lessons, which Jane Willis (1996:17 ) explains as a typical three-part exchange such as that in the example below from our data. (normal rules of punctuation are not applied to transcripts).

1 teacher : right, simge what do you think
2 simge : I..er..er..I think it's correct
3 teacher : yes, I do too. good

In this extract the teacher initiates, simge (a student) responds, and the teacher provides feedback. The teacher gets two "turns" and the student one "turn".

Willis(1996 ibid ) continues:

"In other words it is the teacher who controls the openings and closings of every classroom activity, who controls the turn-taking and who initiates almost every exchange."

The data above reflects, as Willis continues, that : "most of the opportunities for language use are taken by the teacher". Ellis (1990:17) concludes that as regards oral development, "practice makes perfect; learners are more likely to develop in L2 by producing the L2 frequently and correctly". Yet how can learners succeed in doing so when L2 research, quoted in Chaudron, (1988:51), demonstrates that TTT in a typical lesson is, on average 66%, and may at times be as high as 90%? In such circumstances the classroom power relationship is far too highly asymmetrical, ie. it’s too one-sided.

Willis (1996:18 ) calculates that over a 36 week school year, among a class of 30 students, the average student will get only one and a quarter hours’ speaking time. Considering all this, need we be so surprised at our learners' lack of fluency? Need we also be surprised that our students do not know how to initiate or close an interaction, when as the above data shows, all they generally get the chance to do is to respond?

To complement our analysis students completed a survey we had drawn up to investigate learner self-perceptions in 7A, 7B and 7C. The majority expressed general dissatisfaction with their oral ability. They also stated that they'd appreciate the chance to engage in pair work which would :

• give them more opportunities to practise their English.
• ensure that conversations were more symmetrical in terms of classroom power.

The next stage was to gradually introduce our three classes to a variety of pair work and group work activities involving topics and tasks to help develop oral fluency. Involving exchanging information and opinions, these tasks were taken from Ur (1996: module 9) and Willis 1996, who helpfully lists six types of task on page 26/27.

Introducing task-based learning, however, was not easy. The new-found freedom for learners soon developed into noise as "hidden agendas" appeared. To help overcome this, two language monitors ( learners appointed by the teacher) patrolled the classroom during oral activities. Their duty was to admonish / encourage non-participants, and to limit (but not totally outlaw) mother-tongue use. Teachers are perhaps too quick to condemn L1 use, although Eldridge (1997) has pointed out, in his study of 12 year olds in Turkey:

".....in circumstances when the mother-tongue was totally banned in group talk, the resulting interaction tended to be shorter, more stilted and less natural". In other words: less L1 does not mean more L2. The use of L1 often helps the activity run more smoothly and last longer.

The length of pair work activities was slowly increased from ninety seconds up to almost ten minutes. A turn is a chance to speak; a filler is a phrase that precedes what the speaker is about to say. Learners, we noticed, very often lost their "turn". This was because of not knowing how to give themselves time to think. We drew up a list of fillers which enabled them to hold onto their turn whilst preparing what to say, e.g. "I know what you mean"; "I agree"; "wait a moment."

The outcome of our work was that many Intermediate learners who had said nothing (perhaps all term) began to feel less inhibited once it had been pointed out to them how and why they should speak to their partners in English.

The aim of the article above is to outline how attendance at teacher-development sessions may be squeezed into even the fullest of teaching timetables. Our own sessions, for example, enabled us to articulate and begin to overcome a local issue by way of exploratory teaching. Audio-recordings and subsequent transcript-analysis led us to locate possible reasons for limitations on our learners' oral fluency. Improving oral ability cannot, however, occur in only a week or two. A part of each week needs to be devoted to suitable pair-work tasks which reduce TTT, and increase STT in the L2, and which at the same time require learners to negotiate their way through real conversations.

Allwright. R. and Bailey K. 1991: An Introduction to Classroom Research for Teachers. CUP.

Chaudron. C. 1988: Second Language Classrooms. CUP.

Eldridge. J. 1997: Code-Switching in a Turkish Secondary School. ELTJ. Vol. 51/4.

Ellis. R. 1990: Instructed Second Language Acquisition. Blackwell.

Haznedar. A. and Özdeniz. D. 1997 (February): British Council Teachers' Centre Newsletter.

Schon.D.1983: The Reflective Practitioner: How Professionals Think in Action. Temple Smith

Ur. P. 1996: A Course in Language Teaching. CUP.

Wajnryb. R. 1992: Classroom Observation Tasks. CUP.

Willis. J. 1996: A Framework for Task-Based Learning. Longman.