Imagine the following. In a UK secondary school, an English teacher has just finished reading his class of fifteen-year-olds a passage from John Steinbeck's Of Mice and Men. Some students have been handed a simplified edition that trims the language down, yet even that is too much for three of them, and one girl, unable to settle at all, is given paper and told she may draw. From the simplified-edition group the teacher can ask little more than the names of the main characters and a rough outline of the plot. Most of the class he can question about how the characters develop, and around five will back up what they say by pointing to the text. Two especially engaged students, meanwhile, have begun to ask whether Steinbeck means that dreams keep people going or that they end up destroying them.
Speaking as a teacher, I would count a lesson that unfolded like that as a fairly successful one. And yet, wouldn't a class of this sort run more smoothly if the distance between the strongest and the weakest were not so vast? What if every pupil who needed help with literacy were gathered into one room, and everyone eager to debate whether dreams sustain or ruin us into another?
Streaming, or tracking as it is also called, means dividing pupils into classes according to the level of attainment they are judged to have reached. On the largest scale it produces academically selective schools for the most able and comprehensive schools for everyone else. Inside a school it means sorting pupils into a general-ability stream, or into subject-by-subject sets. To almost everyone with a stake in the matter, the idea feels instinctively right. I have heard the mixed-ability approach attacked with a comparison: a group setting off on a hike. The fittest push to the front and set a smart pace, only to halt and wait every twenty minutes; it is maddening, and their eagerness drains away. The slowest, meanwhile, are both mortified and physically at their limit, and the pauses are never long enough. What they secretly want is to give up. Walking, they conclude, is simply not for them.
Classes of mixed ability, so the argument runs, leave pupils bored, parents exasperated and teachers worn out. The brightest will never stand on the peak of Mont Blanc, while the stragglers miss the gentle walk in the park that might have suited them better. The individual is sacrificed to the group, and mediocrity wins out. So: is learning really like hiking?
The prevailing way of thinking about teaching is probably guided-construction theory, which grew out of the work of the psychologist Tomas Varga. In the 1930s, Varga stressed the importance of aiming at each pupil's particular zone of assisted growth, or ZAG, the space between what a learner can manage only with help (from teachers, textbooks, worked examples, parents and the like) and what they can manage alone. Teaching, on this view, exists to supply that support and then withdraw it by degrees, this propping, until the learner stands unaided. Accept the theory, and grouping pupils whose ZAGs are alike seems only logical, while marching everyone up the same slope, whatever their fitness, looks like folly.
For all that, the hard evidence that streaming delivers better results is thin. Professor Ingrid Solberg, who heads the Bergen Institute for Educational Research, remarks that tracking has minimal effects on learning outcomes. Worse still, streaming seems to do real harm to the pupils dropped into the bottom sets, and those pupils are disproportionately drawn from low-income families. Set against that, the modest gain enjoyed by the fortunate few in the top sets counts for little. The upshot is that the able remain able while the rest fall further back, and the gap across society is driven deeper.
In the most recent revision of Solberg's much-cited meta-analysis of what shapes pupil achievement, one of the weightiest factors turns out to be the teacher's own estimate of what a pupil can do. Sort children by measured attainment and you automatically cap how far a teacher believes they can go; keep the room mixed, and those estimates must stay broader and more open.
Streaming may look like a neat way of hitting each pupil's ZAG, but it plays down how much pupils learn from one another. One key strand of guided-construction theory is the part played in building knowledge by the more-capable companion, or MCC. The teacher is the classroom's usual MCC, yet the worth of a well-informed fellow pupil ought not to be dismissed.
It never ceases to impress me how students put an idea across to their classmates in ways that would never have occurred to me. The language and the social signals they use are not a teacher's, and, having grasped the material only moments earlier, they share much the same mental framework as the classmates who are still wrestling with it. There is, besides, a real thrill in handing on something you have only just conquered yourself, a pride, an energy, a freshness in the exchange that the expert, to whom every step is obvious and the delight of discovery long faded, can no longer supply.
A spread of abilities working together in one room is, in short, a rich store of help for meeting everyone's learning needs, and it sharpens communication and social skills too. Now more than ever, it is the many we need to see thrive, rather than have them go under so a handful of shining stars can rise. Once a year I take my mixed class out on a hike. It is hard going. The fittest find they must coax the reluctant along; some scout ahead and report back, and others carry extra loads. And we get there, together.
