Time for a rethink

I think that we can all agree that learning a language is a complex process, involving complex systems. By complex systems, I mean a process of complex behavior which emerges from a few simple rules. All complex systems are networks of many interdependent parts which interact according to those rules. It is a process which is neither linear, nor incremental.
This contrasts with the input hypothesis theory, which has been the prevalent paradigm over the last twenty years or so. The basic premise was that input would be processed, and hypotheses made, which would then result in output where the hypotheses could be tested. This process was said to be innate, and relatively fixed.
If it is the case that language learning involves complex networks and processes, then I think we can safely assume that there is not necessarily one definitive ‘best’ way to learn a language. Indeed, no definitive proof has ever been provided to show that one particular approach is more effective than another.
Under such circumstances, I think we can also agree that teaching is also a highly complex process, involving a multitude of interacting sub-systems. It therefore sometimes makes me wonder why we endeavor to train teachers using an input hypothesis based model where trainers provide the input in the hope that the desired output will be achieved. It is as if training teachers were as simple as programming a computer.
I am going to limit my discussion to the CELTA course, primarily because this is the most established and internationally recognized teacher training course in the world, and because I think this is the one in most need of an overhaul. However, the suggestions I make below could be applied to any pre-service or in-service course.
Although the CELTA was initially conceived as a pre-teaching course, the reality these days is that a great number of people who take the CELTA are not ‘novice’ teachers at all, but teachers with a great deal of experience. The CELTA was also devised at a time when the ‘input hypothesis’ paradigm was the dominant theory behind language learning, and it was also a time when the presentation-practice-production (PPP) approach reigned supreme, fitting, as it did, the input hypothesis model so nicely.
So, I am going to suggest some changes to CELTA courses in the hope that these will better reflect our current understanding of how languages are learnt and the complex nature of teaching. These proposals are in no particular order of preference, and in no way is the list comprehensive. It may also well be the case that some are totally impractical. Some of the ideas will have already been put into practice. Indeed, I know that they have been, with varying degrees of success.

Scrap the term ‘input sessions’

We are dealing with human beings here, not rows of desk top computers. ‘Input’ implies providing data to be processed, with a predetermined output. It reinforces the prescriptive nature of the CELTA course, and the notion that teaching is a linear process.
However, as of yet, I have not managed to come up with a suitable alternative.

Forefront sessions about different approaches at the beginning of the course

There seems to be an unspoken assumption that trainees can only deal with a PPP approach at the beginning of a course, and that more ‘exotic’ approaches will be too challenging for them. By exposing trainees to a variety of approaches right at the beginning of a course, trainees will be better able to select an approach which is appropriate to their aims and their students’ needs.

Make sessions retroactive

What usually happens on most courses is that a session is given and then the trainees are expected to put the procedures and techniques into practice during teaching. I propose that we do it the other way around. The trainees teach and then the content of following session is based on what happened on the classroom. The sessions in effect become more akin to workshops which more readily address the trainees’ needs.

Ditch the language analysis sessions

I have never been convinced of the efficacy of language analysis sessions (modal verbs, conditionals, aspect versus tense, etc). Not only does it beg the question as to which structures should be covered, but only a very limited number of structures can be covered. Surely time could be better spent on dealing with a more important issue, like giving instructions, for example. Or, even better, incorporate more micro-teaching sessions into the timetable. More importantly, trainees should be analyzing the language they are going to teach at the planning stage.

Throw out the Language Related Tasks assignment

In my opinion, this has always been a huge waste of time, and is really boring to do, and mark.
We are basically getting the trainees to spend hours on analyzing language which is not always relevant, when they could be spending the time planning to analyse the language they are going to teach.
I suggest that an assignment based on action research of an area of self-development would be much more useful.

Video all Teaching Practice lessons

Given the technology at our disposal, we should make the videoing of all teaching practice compulsory.
The benefits in terms of development are enormous and it is surprising that we do not make more of this technology. Especially when we consider the complexity of the classroom.
It will enable the trainees to reflect more deeply and accurately on their lessons, unlike in the pressure cooker like atmosphere of immediate post-lesson feedback. Who knows, it might also result in more insightful Lessons from the Classroom assignments.

Lesson plans should not be graded

I think plans should be submitted but not graded. There is a common fallacy that a lesson is only as good as the plan. I think a lot of trainees labor under the misconception that if they spend half the night planning, the more likely they are to deliver a successful lesson. This is blatantly not the case.
If the lesson is successful then does it really matter what the plan looks like? If it is not successful, then maybe it was due to a lack of planning, in which case the trainee will have to plan more thoroughly next time.
The important thing to get across is that the teacher needs to plan in a way which helps him or her give a successful lesson. If that means 20 pages of A4 paper then so be it. If it means a plan written on the back of a box of matches, then that works too.

Get trainees to lead post-lesson feedback as a matter of course

I think we sometimes underestimate our trainees’ abilities to reflect on their lessons and identify areas of strengths and weaknesses, even though some of these trainees may have a number of years of experience.
Of course, a tutor needs to be on hand to prompt and keep the discussion on track. However, we should be seeking to encourage the trainees not to be reliant on the tutor, as if he or she were the all-seeing font of all knowledge.                                                                                                            If we want our trainees to be able to reflect adequately on their lessons by the end of the course, we have got to start promoting this from week one, rather than week three or four, as is usually the case.

Do away with Teaching Practice points

A number of courses do not provide Teaching Practice points. However, some do. I think that the CELTA guidelines should officially prohibit their application. There are good reasons for this. Firstly, the points are usually based on a coursebook and bear no relation to the learners being taught. Secondly, these points are highly prescriptive and actually handicap trainees’ ability to plan effective lessons.

Get trainees to plan the course from day one

The planning of a series of lessons is usually left until the last week of the course. However, there is no reason I can see why trainees cannot be expected to do so from the start. Providing they are given the opportunity to observe the learners before teaching them, of course.

Grade classroom language

The language level of trainees is becoming more and more of an issue as CELTA expands internationally and an increasing number of CELTees are entering the market. The need for accurate language is part and parcel of the complexities of teaching and trainees need to be made aware of its importance. Just as we grade the effectiveness of the lessons in terms of aims being met, we should also grade the accuracy of the language. Explicit reference to a trainee’s language should also be made in the  final report.