In language teaching and teacher development, input is king. i+1 has become a mantra for language teachers trying to enable their students to move from where they are to where they want to be. Short-term, intensive training courses are still the preferred mode of teacher development, despite overwhelming evidence that they are very expensive, have little impact and depend on new ideas and activities presented and demonstrated by the trainer. These ‘new ideas’ tend not to be new at all. The most recent large methodology shift, CLIL (Content Language Integrated Learning) is now 20 years old, CLT (Communicative Language Teaching) over 50. But what happens after the input? That depends to a large extent on the context and the level of commitment of the funding bodies and stakeholders involved in making the change sustainable.

Sustainability in language education means finding ways of maintaining individual development at desirable rates, to prevent the human and material resources within our educational environments from becoming depleted. This involves feeding the system with enough fuel in the form of information, ideas and funding to enable the maintenance of current levels of development or preferably, steady rates of change. Abandonment of this fuelling process, as with almost any dynamic system, leads to disintegration.

A systems approach to educational development enables us to see how different parts of the system impact on each other: where certain parts might need more fuel; where others are being over-stimulated; where feedback loops are becoming fossilised; and what avenues we might take to divert energy flows between different parts of the system to make it work more efficiently, or effectively. Our image of this system appears in Figure 1.

The background environment for the learning ecosystem can also be called the learning context. This includes geographic location, school location within that, education system processes, and the policy environment in the widest view, but also for an individual, their family background, socio-economic situation, and home culture. These factors all interact in multiple complex ways to influence teachers and learners alike.

Within the actual education system itself, the assessment sub-system is key. If it isn’t tested, don’t expect the learners to learn it, teachers to teach it, or parents to care about it, no matter how much training and support the teachers receive. Similarly, enlightened models of education feature concepts such as whole child development, rounded education and supplementary programmes. These require large amounts of energy to maintain, and so are best embedded within the wider educational approach. Without such integrated design, these programmes are likely to be left by the wayside in favour of the curriculum goals that are assessed and graded. Classroom methodologies that maximise these self-sustaining elements, building them in so that they generate intrinsic motivation, making participation in these activities particularly fun and enjoyable, or ensuring some other form of phatic, or actual reward for participation, are the most likely to generate impact.

All learners need support. Teachers are often held solely responsible for this, but what happens at home and in the community is also important. The quality, quantity, and availability of learning materials certainly matters, but the way that learning is managed around these is more important. Availability, quantity, and quality are not sufficient conditions for learning. Individual learners may have the capacity for autonomous learning, but unless they operationalise that capacity, no amount of high-quality materials in their own bedrooms are going to help them develop their language ability. Some people seem to have a natural ability to pick up languages effortlessly. However, they do make a lot of effort, they just know how to organise their own learning more effectively than others.

Organising learning is the main job of the teacher. Knowing what resources can be used to support learners; providing advice to learners and suggested activities to help them make productive use of their time; scaffolding the linguistic content of topics of learning; and giving feedback on the quality of language they are producing: these are the skills teachers need to optimise learning. These skills can be introduced during training courses, but they cannot be mastered over short periods of time. Ukraine has recognised this in its PRESETT English language teacher preparation programme. By increasing the number of practical teaching hours for pre-service teachers from 150 to 650, and drip feeding target teaching skills over a three-year period, they have given novice teachers the time to be able to develop the skills they need to become highly skilled practitioners.

In-service teachers need a support system, too. In Algeria, we have been building that system, over the last four years, with the Ministry of National Education and the British Council, training school inspectors to work together with teachers and school heads to enable deeper change within teaching practice across the wider system. Deep change like this requires deep organisation of the change process and a significant period of time within which to germinate and grow. This project promises to continue for at least another two years.

Sadly, this long-term vision is not the way most education ministries and development agencies operate. Constrained by short term appointments and year to year budgets, the default setting is to run training courses for as many teachers as possible, and equate the number of people taking a course with its impact.

The number of people taking a training course is NOT impact. Depth of change within the system in terms of quality of teaching behaviour and consequent increases in quality and quantity of learner outcome achievement are the measures by which education change programmes should be judged. That involves a lot more detailed monitoring and evaluation than head counts or checklists. Monitoring and evaluation systems, when well-engineered, should be a component of the project structure that helps enable change processes to proceed.

As all complex systems are sensitive to initial conditions, small fluctuations in any one of the inputs into the system (materials, teaching/learning, assessment, teacher support, infrastructure, ideas and information) can lead to unpredictable effects on large-scale outputs. For example, in West Africa, our Toolkit for English Clubs is part of the learning support system that is engendering changes in classroom teaching. At the moment, we are witnessing the potential for that small intervention to have a huge influence over classroom teaching across the secondary system in some of the countries where it is being used.

Building sustainability into our change projects involves considering the people involved and the processes they are taking part in. We need to take a human perspective on the actual changes we are asking individuals to make, the real effect they are likely to have on the participants, and how different these new ways of working are from those already embedded and engrained. Appreciating the traditions behind current practice respects existing practices and helps empathise with individuals we work with. Ensuring that our project partners believe we are suggesting changes that have their best interests at heart is more likely to ensure buy-in from the outset, and support deep change in the future.