I am delighted to be here this afternoon to talk to this audience. My talk this afternoon proceeds in three stages.
- The problem of generic approaches to teacher education, where generic means ‘not subject-specific’
- How we might rethink teaching as a subject-specific practice.
- What this looks like in the context of preparing mentors.
In this talk I am going to be drawing on a number of sources, including my own doctoral research into the subject-specific nature of history teacher discourse, my experiences as a history mentor in the Cambridge partnership, and more recently my experiences of working with novice and experienced teachers in a multi-academy trust.
My background is in mentoring and training trainees in secondary history, although a lot of my recent work has been with primary school teachers, and, although I would want to exercise a little more restraint and caution in my conclusions when discussing primary, I think the majority of the points I want to make this afternoon apply in training primary teachers as secondary.
As we will all know, Shulman’s work in the 1980s proved influential in establishing a research tradition in teacher education in which the knowledge base of teachers was theorised. His work, rightly, helped us to focus on what it is that a teacher knows, and not just on what a teacher does.
For those who read the blogs, this idea that a rich knowledge base always sits behind skilled performance is increasingly a dominant position in our educational discourse. But Shulman pre-empted this by decades in arguing that, if we want to look at teacher quality, we need to look at the knowledge base that sits behind skilled performance. His initial categories – content knowledge, curricular knowledge, and pedagogical content knowledge – helped us make sense of this knowledge base.
But whilst categories can be helpful, they can also be distracting.
Shulman himself developed his model in more complex ways, as did his students, and countless subsequent writers who have sought to theorise the knowledge base of teachers. The categories have proliferated, and the status of different categories has in some cases been confused. By way of example, a brief search online for how the term ‘pedagogical content knowledge’ is used will offer a rich variety of definitions, some of which incorporate things that Shulman himself explicitly excluded in his original categorisation.
In one of my more mischievous moods, I might suggest that terms such as ‘pedagogical content knowledge’ have come to mean so much that they now mean very little. The point of a category is to help create shared meaning, but if a wide array of meanings are used to define a category, then we lose the very purpose of creating that category in the first place.
I would gently suggest that we probably spend too much time theorising the categories, and not enough time theorising what sits within the categories.
This matters to us as teacher educators, of course, because what constitutes that knowledge base is what we need to teach our trainee teachers.
And, as we all know, the person who does the lion’s share of the teaching is not you or me or some speaker at the Summer Institute, but rather the mentor on the ground, in the school, with whom the trainee spends most of his or her time.
It is for this reason that I follow Christine Counsell’s imperative, which is that the most important part of our job is not to train trainees, but to train mentors. If we want to ensure that trainees are well-trained, then the single most important thing we can do it to ensure that the mentors are well trained. To put it another way, the knowledge-base of the mentor should be our responsibility.
Where we do not make this our priority, bad things follow. One possibility is that there is a disconnect between what is taught in universities and what is taught by the mentor. This results in trainees failing to see the relevance of what is taught outside of school. Another, worse, possibility, is that the mentor and the university tutor teach conflicting things. How confident are we that mentors in schools do not issue the fatal line of “don’t worry about what they teach you in university – I’ll teach you what you really need to know”.
As bad as this sounds, I think there is however a worse problem if we do not attend to the knowledge base of mentors, and this is that the tools of the final summative assessment become the curriculum for trainees.
Everyone here will I am sure have read Christodoulou’s recent work on assessment and progression models. In it she offers a withering critique of what happens when summative assessment frameworks – such as the National Curriculum level descriptions, or GCSE marking criteria – become the curriculum that is taught. This is all the worse, she argued, when those criteria are broad, vague and generic.
And, although Christodoulou’s work focuses on the teaching of children, exactly the same point can be made in teacher education, and it is one that Shulman made all those years ago in his 1986 article. In this he railed against generic assessment in teacher education that focuses on the practices of the teachers rather than the knowledge base that sat behind this. He wrote
“In most states, however, the evaluation of teachers emphasises the assessment of capacity to teach. Such assessment is usually claimed to rest on a “research-based” conception of teacher effectiveness. I shall take as my example a list of such competencies prepared by a state that I briefly advised during its planning for a state-wide system of teacher evaluation. The following categories for teacher review and evaluation were proposed:
1.Organisation in preparing and presenting instructional plans
3.Recognition of individual differences
7.Educational policies and procedures
As we compare these categories…to those of 1875, the contrast is striking. Where did the subject matter go? What happened to the content?”
This was in 1986, and Shulman’s work is some of the most cited in the field of education. So after 30 years, where have we got to?
Well, we have the Teachers’ Standards. And, although subject knowledge is dropped in to a number of the standards, no where is it set out what knowledge a history teacher, or a teacher of Year 3, needs to know.
Decades of research on Shulman’s categories has fundamentally failed to resolve the very problem that Shulman identified. We have even managed to turn teacher knowledge into a competence.
All of this then collided in a perfect storm that mirrors exactly what happened in the teaching of children in the late 1990s and 2000s. The ingredients are
- generic and vague statements of competence designed to be used summatively
- an absence of specific curriculum that sets out what knowledge base ought to be learnt
- stringent accountability measures based on summative data
In these circumstances, as Christodoulou has shown so well in the context of school assessment, the generic summative competences get used as the curriculum. We end up in a situation where trainees are set targets using the wording of the Teachers Standards. We end up in a situation where institutions and organisations try to apply Ofsted-style grades to these generic competences, and then track trainee progress in terms of whether those grades improve.
And it gets worse. When we set targets for trainees, we should be setting our sights on a gold standard, what we aspire for our trainees to become. When the Teachers’ Standards are used as targets, the minimum standard becomes the gold standard. This is almost the definition of low expectations. It is not difficult to see how some mentors and trainees can see the whole training process as a box-ticking exercise.
So much for the state of things at the moment. What can we – what ought we – to do about it? I think there are two things we need to do. First, I think the answer lies in going back to Shulman’s original imperative, and to ask ourselves what knowledge sits behind the skilled performances of teachers. Secondly, I think we need to make it our jobs to ensure that the mentors themselves have the knowledge base so that they can teach it to novice teachers.
In this second part of my talk, I shall be drawing on the work done in the Cambridge partnership, specifically by the community of history mentors under the leadership of Christine Counsell. Over the last fifteen years or so, that community has attempted to do exactly this.
Now, of course, there clearly is some knowledge which is ‘generic’, by which I mean ‘applies to all subjects’. I want my trainees, for example, to know about the limits of working memory or what they need to do in the morning to prepare their vocal cords for a day of teaching. Shulman’s own delightfully defunct example is “break a large piece of chalk before using it to stop it squeaking on the board”.
But the experience of the Cambridge history mentors was that almost everything we wanted our trainees to know was grounded in subject-specificity. So let’s take some examples of the day-to-day practices of teachers, which are often called ‘teaching skills’. These might be things like:
- lesson planning
- giving an explanation
- giving feedback
It is not at all uncommon for these to be understood as generic teaching skills which can be learned in a generic way and then applied in the context of the subject being taught.
But what if we take perhaps the most important and most challenging aspect of teaching, which is to ascertain what the knowledge base is of the people you are teaching. This is often captured in terms like ‘formative assessment’, ‘assessment-for-learning’ and so on, and we are all familiar with websites offering “ten top tips for formative assessment” and alike.
But what is actually going on when you assess someone’s knowledge? In short, it involves a comparison, between the mental map of the subject you have in your head and the mental map they have in their head. In making an assessment, a teacher is trying to diagnose where the gaps in that mental map are, so that he or she can then do something about it.
To make a judgement call on something as complex as this requires that the teacher’s mental map is very strong. Knowledge of common misconceptions might help highlight the parts of the mental map which are most likely to be missing or misunderstood, but knowledge of common misconceptions is an enhancement of, and not a replacement for, that mental map.
In this way, the ability to assess is a product of one’s own subject map. The richer that map of the subject, the more accurately one will be able to identify where the pupil has weaknesses.
Let’s move to questioning, which is often taken to be a generic teaching strategy. We have I am sure all seen mentors write targets such as “improve open-ended questioning” or “make more use of recall questions to enhance memory”, and that sort of thing.
Again, though, a teacher’s ability to question is a product of the mental map the teacher has of the subject.
To question someone about something is a deeply subject-specific act. For starters, one must immediately make a judgement call on what are the questions worth asking. What counts as a ‘closed question’ or an ‘open question’ depends on where that question sits within the map of the subject being taught. If you’ll excuse my Aristotelian leanings, what constitutes ‘good’ depends on the purpose of the act, and the purpose of the question is to explore further the map of the subject. It follows that what is ‘good’ and what is ‘bad’ is dependent on what is being taught, rather than some intrinsic feature of the form of questioning itself.
Time here prevents me offering more examples, and I do not want to be dogmatic on this point – I am quite willing to accept that there are some generic things that a teacher needs to know. But my point would be that our starting point should always be the subject being taught and that, in the vast majority of circumstances, the subject can furnish us with the answers we need.
Our challenge in the Cambridge partnership was therefore to support mentors – these most crucial people in the training process – in knowing what they needed to know to be able to put the subject first. My sense of this was that five crucial elements were needed, which underpinned everything that took place in the course.
- Targets had to be recognisable as being for the teaching of that subjects.
- Comments in written observations had to be tied to a map of the subject being taught.
- A training activity designed to meet a target should naturally point towards some subject-specific literature that a trainee could read.
- Written reports needed be dripping in references to the subject being taught.
- The culture of the mentor community should be one of subject scholarly expertise, with the mentors taking ownership for the content of the course.
Doing this in the context of a relatively small and geographically focused university partnership or regional SCITT is somewhat easier. Achieving this with a national organisation of the scale of Teach First is somewhat harder.
I remain convinced, however, that the quality of teacher education in this country needs this emphasis on subject-specificity. University tutors have a vital role to play in this process, particularly in terms of training mentors, and curating the knowledge base of the expert mentor.
But I would finish simply by reiterating Shulman’s missive. Teacher education is weak when we focus on generic competences. It is at its strongest where we define and then teach the knowledge base that an experienced teacher needs to have.