ESL TeachingTraining

The Case Against a ‘Clinical Supervision’ Model in Language Teaching

“ the observer observes the teacher teach and collects objective data on that aspect of teaching they agreed upon earlier.” (Hopkins, D. 1985 :88)
“ clinical supervision, which we define as the process by which teaching performance is systematically observed, analysed, and evaluated.” (Gaies, S. and Bowers, R. in Richards, J. and Nunan, D. 1997 :169)


Observation (or clinical supervision) of language teachers has long been recognised as an essential part of continuing professional development (CPD) but in recent times pressure has been applied to teachers to conform to imposed criteria in their teaching which have been designed primarily to enable institutions to qualify their teaching publicly.

The quotes above offer an insight over 12 years into the development of the quasi-science of ‘clinical observation’ which now permeates the world of language teacher-training and development. Arguably, one of the causes of this evolution is the drive to enhance the professional image of the industry - but some may think that professionalisation does not need to go hand in glove with falsely premised measurability.

This essay will attempt to offer an alternative view in contrast with the commonly accepted notions that observation, to be effective is necessarily:

Observation of teachers has little direct connection to student outcomes (i.e. learning)


I shall begin by questioning the view that what a teacher does in the language classroom, and therefore that which can be observed using accepted observation techniques, has a direct bearing on the language learning outcomes of students. It may seem absurd to challenge this view and yet in language education we have strong indications on a daily basis that teaching does not equate to learning (I shall avoid the controversy of implying that this may well be true in other disciplines).

This challenge to received wisdom may seem especially incongruous in the light of UK government initiatives to precisely quantify language learning and teaching as embodied in the ESOL syllabus and the bureaucracy attached to it. Is seems that if we tick enough boxes and provide enough meaningless comments, we can be assured that the system works.

However, Ellis’ comment that

“..what is learned is controlled by the learner and not the teacher, not the text books not the syllabus.” (Ellis, R. 1993 :4)

clearly states the case that every experienced language teacher knows to be true. Students cannot be ‘learned’ through teachers’ input since input does not equal output. Language learning is a series of procedural events some of which are controlled by exposure, some by instruction but most by internal psychological and cerebral mechanisms over which teachers have little or no control (and arguably should not have).

Having said this, researchers and teachers in the field tend to agree on at least one point in regard to how classrooms should be organised and run – that point is affectivity.

A small sample of the literature supports the notion that teachers should be attending to the affective aspects of teaching and learning:

As early as 1981, Piepho’s advice is crystal clear:

“the crucial factors which contribute towards successful teaching are the overall ‘pedagogic atmosphere’ and in particular the personal relationship between teacher and learner, learner and learner.”(Piepho, H-E. in Candlin, C. 1981 :10)

“Emotions must [therefore] be considered an integral part of learning..”(Williams, M. and Burden, R.1997 :28)

“The affective domain includes many factors: empathy, self-esteem, extroversion, inhibition, imitation, anxiety, attitudes – the list could go on. Some of these may seem at first rather far removed from language learning, but when we consider the pervasive nature of language, any affective factor can conceivably be relevant to second language learning.”(Brown, H.D. 2000 : 64)

“The affective variable: anxiety needs to be lowered for learning to take place.” (Harmer, J. 2001 :96)

Richards too notes that experienced teachers, perhaps because they are able to be more flexible, adjust their lessons for interest and motivation. (Richards, J. 1998 : 115)

There seems to be common agreement at all levels that classrooms in which students are valued, feel comfortable and unthreatened and are able to contribute freely are likely to foster learning. Clearly the teacher is a central component of the creation of such an atmosphere, assuming that there is also institutional support for such an approach.

If we accept that a major contributory factor to efficient and effective learning is the affective environment created in large part by the teacher and her interactions with students, then we must ensure that observation is focussed on the fostering of such an environment rather than on the content delivery (how a particular tense, functional chunk or lexical item is presented, for example). This then takes the focus away from teacher performance and student outcomes, instead nurturing the conditions in which learning may take place.

Observation should not evaluate teachers but offer support for further development.


Most teachers, even experienced ones, object to being observed. The anecdotal evidence for this is partly in the fact that it is considered ‘polite’ to pre-warn or request permission to observe a teacher at work. This seems superficially to be perfectly reasonable and is widely accepted as the norm – in fact the principle has rarely (if ever) been questioned in the literature.

But why should a classroom become a closed and secret place? Surely, an adult language classroom should be accessible to (almost) all; colleagues, line-managers and even the interested public!

But Roberts makes the point clearly that observers disrupt lessons:

“Routine teacher behaviour and learner performance are rarely seen during a one off observation, as it disrupts normal class activity”. (Roberts, J 1998 :168)

And Hopkins reminds us that 20 years ago we felt the same way:

“Many teachers feel distracted by the presence of an observer and feel that despite the intention that observers should observe and not evaluate, it is impossible to watch another teacher without forming some sort of impression of how successful the lesson was.” (Hopkins, D. 1985 : 142)

Perhaps then there is a sense that teachers need time to prepare a ‘special lesson’ because there is an outside body present. This, it seems to me, is akin to bringing out the best tablecloth to create a ‘good impression’. We are left then with something of a conundrum: if lessons are specially prepared, the observer is not seeing a true picture of events and if a representative lesson is not being observed, why observe it?

Lesson observation has been used traditionally for evaluative purposes rather than for genuine Continuing Professional Development (CPD) motives. Teachers feel ‘watched’ rather than supported and this is reflected in advice given to teacher-trainers and observers in general, to sit at the back, not interfere and not take part for fear of disturbing the teacher! For example:

“The observer should remain an observer. An observer cannot observe effectively if participating in the lesson.” (Richards, J. 1998 :143)

This comment seems undeniable and yet if one re-examines the purpose of observation and concludes that the overall gaol is to support the teacher and maximise learning opportunities in a given context, one is left wondering if perhaps a participating observer might not also bring much to a later discussion.

While I am not arguing for random team-teaching or an interventionist style of observation (at least not as a norm) I suggest that two aspects of teacher observation need attention:

  1. supportive observation should be casual and reciprocal
  2. feedback should be a dialogue between interested peers

So 1, above, suggests that if teachers can habitually ‘pop in’ on each others’ lessons then the notion of the special event can be diminished. Furthermore, the important matter of trust can be addressed. If teachers are visiting each others’ classes as a norm, then they may begin to accept the non-judgemental status of such visits.

2 relates to a post-lesson feedback session. Although this essential aspect of observation is not under scrutiny here, it is nevertheless worth noting that if teachers and observers have shared an experience, rather than one being subjected to the other, subsequent discussions will be more equally balanced.

Woodward makes a relevant point:

“If observers don’t bother to try and grasp the teacher’s logic or to sense what the teachers and students are sharing, then they are just gatecrashers at someone else’s party – unwanted trouble-makers at an otherwise pleasant event.” (Woodward, T. 1991 :66)

The observer then, needs to engage with the observee in understanding the context and events of the lesson or lesson segment to create an exploratory discussion rather than a list of ‘positive’ and ‘negative’ points.

Teachers who have been observed need to be able to discuss the event with a partner not defend themselves against an onslaught.

These simple features of observation can help to reinstate it as a process designed to support rather than to judge.

But we are battling a short tradition in which the observer is viewed less as an agent of developmental process and more as a ‘spotter of errors’ as is starkly evidenced here:

“The training role of the supervisor is fairly obvious: it involves the use of classroom observation to identify deficiencies and to bring these to the teacher’s attention.” (Gaies, S. and Bowers, R. in Richards, J. and Nunan, D. 1997 :169).

Observation cannot be objective


Anyone who has observed a lesson as a trainer involved in deciding whether it (the lesson) should pass or fail will recognise the uncomfortable sense that a different trainer might reach a different conclusion. We know, but try to suppress the thought, that our observations cannot be objective. As Fanselow (1997) notes: when different people discuss the same lesson, they often give contradictory accounts of the same event.

If one holds the view that language teaching is a science, then one can construct a simple procedure for teachers to follow and learning will take place. Attached to the procedure, we can create a pro-forma which will allow any minimally trained observer to take appropriate notes for the purpose of postobservation feedback. In fact there are many such quantitative systems which, for example, allow the observer to note the frequency of certain events; examples would be Flanders Sign System, the Communication Orientation of Language Teaching (COLT), Target Language Observation Scheme (TALOS) and Fanselow’s Foci on Communication Used in Settings (FOCUS).

Since attempts to create a rational and science-based system for language teaching (Cf Audiolingualism) have never met with enduring success, one must ask how a systematised observation strategy could possibly be applied to a discipline which is a variable combination of science, art, psychology, social-interaction, cultural exchange and more.

In addition, it is now commonly agreed that the SLA research community are not yet able, and perhaps never will be able, to describe and predict exactly how language learning takes place. Some teachers can provide anecdotal evidence of how their long-term students have performed but, interesting though these snippets may be, even an accumulation of such fragments cannot be relied upon as ‘scientific evidence’.

We are forced then to agree that as a profession, we know little about how learning takes place and so must be engaged in educated guess work as far as teaching is concerned. It follows that we cannot create an entirely objective context for observation to take place – we are involved with a fluid, creative combination of teacher and students reacting together in an event which will never be repeated - a lesson is a series of unique moments and should be treated as such.

An observer to these unique moments should be equipped to observe them not bound to tick boxes which relate to some ideal lesson in which Stage 1 precedes Stage 2 and can never be juxtaposed. Observers should be prepared to sense the flow of the lesson, how students interact, whether they are willing, able and have opportunities to create and use language. Attention can still be paid to ‘problem moments’ (students who don’t understand something, object to something, lose concentration etc) but since we have all experienced them, we must also recognise that they are a natural and expected part of the craft of teaching. Most importantly, the observer should be able to gauge the affective environment, the mood, atmosphere and ambience.

This notion of observation cannot be described as objective but must, by the nature of the event, be largely subjective. I would like to introduce the phrase: informed subjectivity, by this I intend that a well prepared observer (who can as well be a colleague as a trainer) and the teacher being observed have agreed (either specifically or tacitly) on a variety of approach-based ‘givens’. These ‘givens’ form the basic working model which is used within the profession, in that institution or even, at a micro level, in that particular classroom with those students. As Richards notes:

“.. the significance of what is observed depends on the theory of teaching (or the particular approach or methodology) the teacher holds.” (Richards, J. 1998 :142)

and to this I would add: and the specificity of the moment.

An example may help to clarify: a teacher may be working on a particular problem of group dynamic during the observed lesson segment (perhaps because students have not been taking full advantage of available opportunities to interact) and unless the observer understands this, could not be expected to offer relevant and pertinent comments – one might have expected a higher language focus and may not have the insight to understand the long-term context.

Informed subjectivity means that both observer and teacher can focus their attention during feedback on the variable processes which, while unique to the moment, accumulate to inform a teacher’s experience and therefore contribute to CPD.

Observation cannot be systematic


In a sense, the argument presented above can also be applied here but may require additional representation.

Systematicity suggests order and sequentiality, two features evidently judged by teacher-training exam boards (UCLES and Trinity College London being the most influential) as essential to a ‘good lesson’. Course participants are required to present detailed lesson plans to the observer prior to the lesson and then to follow those lesson plans. Lesson plans are then scrutinised by exam-boards representatives apparently in the belief that they reflect a trainee teacher’s ability to teach. This insistence on prelesson organisational ability may offer insights into skills of prediction or logic but do not in any way connect with a classroom full of people and one of their number’s capacity to create a learning environment.

It seems to me that observation which is linked to a lesson plan sequence, does little to enable either teacher or students. In fact it appears to be testing the ability of the teacher to follow a plan – perhaps this is an admirable trait for a teacher but I am not aware of any evidence to the effect that pre-ordered lessons create better learners. Richards makes the point that:

“The mechanist model, with its assumption that planning equals teaching equals learning simply does not match the reality which is that planning, teaching and learning are complex, multidimensional activities..” (Richards, J. 1996 :44)

I am not arguing here for anarchy in lessons, nor am I suggesting that lesson plans are redundant; in fact, order, perceived as such by students and teacher is to be valued. The difference is that in valuing this careful, prepared sequencing so highly, exam boards and some institutions have lost sight of any alternative and set an authoritative precedent to the rest of the profession.

We know that language learning is not, as was once thought, a matter of presenting grammatically sequenced items, one after the other and having students reproduce those items at a given moment. We know that it is a far more holistic and generative process relying not simply on instruction time but exposure, mood and to some extent, natural and learned learner strategies. In our current state of knowledge, this is indisputable so one could argue that if teachers were encouraged to be less systematic and more reactive, we might be able to respond more effectively to the needs of our students.

Observation then should surely be a stylistic match for the activity of its attention – it should be able to wander, to change tack, to focus on a variety of areas in no particular order and it should also be able to follow a clear and ‘logical’ trail if indeed the lesson did. But my point is that observation should be free to find its own direction informed, uniquely by teacher, students and context.

Observation should be almost anything but clinical


Lexical associations for clinical might be; detached, cold, distant or irrefutable and proven. It is clear that trainers (in particular) have been advised to be detached in their observations, to never ‘get involved’ and the rationale has been that one mustn’t disturb the teacher, an argument I have approached above.

Few would encourage an observer to be ‘cold’ since there can be no reason (even on an uninformed basis) for such a position, except perhaps the all too common attempt at self-ingrandisement. Furthermore, only a grossly ignorant observer would argue that conclusions they reach while observing are ‘proven’ or ‘irrefutable’.

There seems then to be no reasonable support for the notion of clinician-observer/supervisor as long as we are in accord that the overarching aims for observation of teaching as a human activity is to enhance the learning environment through encouraging teacher development.


I have attempted to present the case for a less evaluative observation to nurture a more valuable observation experience. To achieve this both observees and observers must be prepared to abandon traditionally held roles and embrace a more humanistic notion of process over product, whole over discrete, environment over technique and learning over teaching.

I have taken a relatively idealistic stance in the face of firm belief systems and axiomatic acceptance; yet some comfort may be taken from the fact that teachers with whom I have spoken all over the world agree that ‘observation’ is ‘useful’. Rather as a medicine can be ‘useful’. This general agreement could be exploited, in the most positive sense of the word, and built upon to create a more meaningful model called, perhaps: Co-operative Observation.


Basic Skills Agency; Steeds, A (Ed)Adult ESOL Core Curriculum2001 DFES
Brown, H. D.Principles of Language Learning and Teaching (4th ed)2000 Longman
Piepho, H-E. in Candlin, C. (Ed)The Communicative Teaching of English1981 Longman
Doff, A.Teach English (Trainer’s Handbook)1988 CUP
Ellis, R.Second Language Acquisition Research: How does it help Teachers?1993 ELTJ 47/1 OUP
Fanselow, J.Breaking Rules1997 Longman
Gronlund, N.Measurement and Evaluation in Education1981 Macmillan
Harmer, J.The Practice of English Language Teaching (3rd ed)2001 Longman
Hopkins, D.A Teacher’s Guide to Classroom Research1985 OUP
Malderez, A. and Bodoczky, C.Mentor Courses1999 CUP
Nunan, D and Lamb, C.The Self Directed Teacher1996 CUP
Richards, J and Nunan, D. (Eds)Second Language Teacher Education1997 CUP
Richards, J.Beyond Training1998 CUP
Roberts, J.Language Teacher Education1998 Arnold
Ur, P.A Course in Language Teaching1996 CUP
Williams, M. and Burden, R.Psychology for Language Teachers1997 CUP
Woodward, T.Models and Metaphors in Language Teacher Training1991 CUP


Feel free to contact me about this article by email:

Learning by List