November, 2020. Afzal Khan, head of science in a specialist college and university clinical lecturer in science pedagogy, arrives at school. Checking his PDA for messages, he reviews the day ahead and fine-tunes plans for his morning work. There is a lead lecture to 100 Year 9s, with follow-up lessons taught by the eight intern teachers on placement in the school. And then at the end of the morning he has a structured lesson-study seminar led by Afzal and Marcus Swain, one of the university’s research academics, using video from the morning to develop the intern teachers’ understanding of pupil misconceptions about genetics.

Over lunch, he joins other teachers from the school-university research partnership, who are working on a development project on equity issues across the curriculum led by Charlotte Smith, a professor in teacher learning from the university and consultant headteacher in the school cluster.

In the afternoon, Charlotte leads a workshop for all the school’s student teachers on researching professional practice, which is video-conferenced across the cluster while Afzal teaches his own post-16 science group.

When the pupils have departed, Afzal, Marcus and Charlotte meet the whole-cluster teacher education RD team for a seminar drawing together findings from current projects. In the early evening, they lead a workshop for parents as part of the cluster’s family-learning initiative.

For us in 2011, much of Afzal’s activity is recognisable - 2020 is not that far away - but the overall coherence of the work is novel. Afzal, Marcus and Charlotte work in a framework that draws together universities, schools and teachers in a research-led, practice-focused approach to professional learning.

The formal roles they occupy are part of a systematic structure that draws together university and schools into a formal relationship focused around teacher learning and development. Heads and senior university staff co-operate in a common governance structure to make it all work.

For far too long, our thinking about teacher education has been shaped by a set of outdated oppositions. We assume a distinction between “theory” and “practice”, but this makes no sense in a profession that is both highly practical and drenched with theory about learning, the social contexts of children’s lives, and the role and nature of education. We assume a distinction between “university-based” and “school-based” training, when patterns of learning and technology mean modes of learning are not dependent on location.

We assume a distinction between “craft-skills” and “knowledge”, when both intertwine at all stages in developing the practice of teaching. We assume a distinction between initial training and post-experience practice that makes no sense in an environment of lifelong professional development.

These distinctions hamper the ability of teacher education to realise its potential in shaping a high-quality profession that provides consistently outstanding learning opportunities for young people. Of course, Ofsted evidence is that the overall standard of teacher education is high. But it could be far better if we seized the opportunity to rethink teacher education as other professions, including law, medicine and engineering, have done.

We need to recognise that 21st-century teaching requires a step-change in the way we prepare, support and develop teachers, and that neither schools nor universities can do this alone.

Since the advent of compulsory education in the 19th century, teaching has too often been seen as a semi-profession based on craft skills. All this despite enormous shifts - accelerating over the past 15 years - in our knowledge about how learning happens and what makes for effective and successful teaching. In the 1950s, much medical training was not dissimilar: medical students learnt their skills through apprenticeship to a consultant - caricatured in the Lancelot Spratt character of the film Doctor in the House - and largely replicated the practices they had been shown.

Since then, medical training has undergone a revolution. Professional training is lifelong. It is informed by research; it integrates field learning, reflection, creativity and innovation, often through problem-based learning methodologies. It has to be like this because of unprecedented advances in medical knowledge and technology.

The same is true of teaching. Huge advances in knowledge about how people learn, the rapid development of technologies, and the pressing need for economies to raise skill and knowledge all mean that it is time to move beyond the vestiges of apprenticeship to models of professional learning.

Current teacher education has strengths. The development of partnership-based training has changed practice considerably. Evidence from new teachers and heads confirms that most teacher education delivered through university-school partnerships is at least good, and often outstanding.

But these partnerships are too often relatively transient and uneven, and are insufficiently embedded in the cultures of either schools or universities to deliver the professional training we need.

Too much teacher education aims to secure classroom competence, rather than seeing classroom competence as a stepping-stone to outstanding professional practices. Some of the world’s most interesting teacher education programmes, such as the Melbourne master of teaching degree programme or the Stanford teacher education programme, are far more rigorous and far more transformative of schools, of universities and of teachers than much current English practice.

In the Melbourne programme, the curriculum of university training and the curriculum of schools are geared around developing new teachers as advanced users of evidence to underpin focused interventions in children’s learning, with university staff embedded in schools.

The next steps in the transformation of teacher education involve some bold thinking, which goes beyond those outdated oppositions that bedevil the subject. The Government’s November white paper, The Importance of Teaching, sets out an aspiration for a school-led system of teacher education, and of course schools need to be central to initial teacher education.

The proposed teaching schools have enormous potential to help develop such ideas, but, as many are realising, even the best schools need to transform themselves to make it work. They need to place adult learning at the core of their practices so that the generation of knowledge about successful teaching is at the heart of their development planning, and that learning to teach better is built into the work of teachers.

Universities are central to this endeavour. It is in universities that research-based knowledge is generated, through testing new practices against rigorous standards of evidence and where practices can be generalised. It is not enough to know what works, because most things “work” in some circumstances. But it is important to know what works best, and why, and how it will be susceptible to changes in curriculum, technology, assessment and changes outside schools, and to get this into the warp and weft of schools.

This will involve new staffing structures, novel patterns of teaching and learning, and imaginative curriculum development. Beyond this, we are unlikely to deliver the sustained, consistent and coherent professional learning for teachers we need simply through local innovation, important though this is.

What is needed is what Stanford University’s Linda Darling-Hammond calls a “teaching and learning system”: an overarching organisational and cultural structure that will nurture and sustain the work of professionals like Charlotte, Marcus and Afzal. And to make this happen, we need to reconstitute relationships between research-led universities and successful schools.

These reconstituted relationships need to be much more deeply embedded than either current partnerships or models based on commercially trading services can be.

One useful way of thinking about the sort of relationships we need is, again, derived from medical education. The training of doctors revolves around a “deanery”, which works across medical institutions to plan, organise and develop medical education. In teacher education, this would be a regional structure that brings together the key players - schools and universities - in a common framework geared around developing research-based teacher practice, recognising that world-class education needs unity of purpose.

We do have good teacher education in current partnerships. Now we need the confidence, imagination and ambition to create structures that will allow us to genuinely transform the teaching profession to meet the extraordinary challenges of learning in the 21st century.

Chris Husbands is director of the Institute of Education, University of London.