Schools as we know them have existed for a very short time; they're a vast social experiment. If the hypothesis is that universal compulsory schooling is the best way to to create an informed and critically literate citizenry, the results are, at best, mixed. Ancient cultures knew to wait to give a child a task until you see that she is ready for it; that if a child tries something and then backs away, you leave him alone, because he will try again later; and that young children learn better from story than lecture, from hands-on experience than direct instruction. Any wildlife biologist knows that an animal in a zoo will not develop normally if the environment is incompatible with the evolved social needs of its species. But we no longer know this about ourselves. We have radically altered our own evolved species behavior by segregating children artificially in same-age peer groups instead of mixed-age communities, by compelling them to be indoors and sedentary for most of the day, by asking them to learn from text-based artificial materials instead of contextualized real-world activities, by dictating arbitrary timetables for learning rather than following the unfolding of a child’s developmental readiness. Common sense should tell us that all of this will have complex and unpredictable results. In fact, it does. While some children seem able to function in this completely artificial environment, really significant numbers of them cannot. Around the world, every day, millions and millions of normal bright healthy children are labelled as failures in ways that damage them for life. And increasingly, those who cannot adapt to the artificial environment of school are diagnosed as brain-disordered and drugged. It is in this context that we set out to research how human beings learn. But collecting data on human learning based on children’s behavior in school is like collecting data on killer whales based on their behavior at Sea World.