Unschooling is the name John Holt gave, in the 1970s, to a practice he had concluded was prior to schooling: children learning what they want, when they want, with adults nearby who help when asked and otherwise stay out of the way. Holt arrived at the position by way of disappointment. He had spent years as a teacher and a reformer, trying to make schools more humane, before deciding that the institution itself was the problem — not its worst examples but its basic shape. His 1964 book How Children Fail described what classrooms do to thinking. His 1967 How Children Learn described what children do when classrooms are absent. By 1976, with Instead of Education, he was advising parents to remove their children entirely and let them learn the way humans had learned for most of human history.

The central claim is empirical, not ideological. Children, given access to interesting adults, interesting materials, and unstructured time, will learn to read, count, reason, build, argue, and concentrate without being made to. They will do it on a schedule that does not match the school's schedule. Some will read at four, some at nine, and the nine-year-old will catch up within a year. Some will spend a season obsessed with frogs and a season obsessed with code and a season doing what looks to outsiders like nothing. The "nothing" is usually the consolidation phase, when the previous learning settles. Unschoolers and their parents become connoisseurs of this rhythm and learn to trust it.

Peter Gray, an evolutionary psychologist who joined the movement intellectually in the 2000s, gave Holt's empirical hunch a deeper grounding. In Free to Learn (2013) and the surrounding research, Gray argued that the human child is a self-educating organism the way the human child is a self-feeding organism — built, by long evolutionary inheritance, to learn through play, observation, and imitation in mixed-age groups. Schooling, on this account, is a recent and odd intervention that works only by suppressing the older system. Remove the suppression, restore the conditions (mixed ages, free play, real responsibility, access to the adult world), and learning resumes its natural shape.

The practical implementation falls along a spectrum. At one end, "radical unschooling" abolishes not only curriculum but bedtimes, screen limits, and food rules — anything imposed by the parent on the child. At the other end, "relaxed homeschooling" looks like a household with a flexible curriculum and a lot of room for the child's interests. In between sits the version most unschooling families actually practice: no formal lessons, but plenty of conversation, books in every room, a couple of standing commitments (a music teacher the child chose, a math program the child agreed to), and frequent contact with other unschooling families. The pure form is rare. The hybrid is durable.

The collective dimension of unschooling is where it gets interesting and where its critics have a real point. A single unschooling family in a neighborhood of school-going children produces a lonely child. A community of twenty unschooling families produces a functional alternative. The early Holt-era movement was scattered. The current movement, organized around hubs like the Alliance for Self-Directed Education, networks of "agile learning centers," Sudbury-model schools, and the loose constellation of homeschool co-ops, has begun to provide the social density that makes the model livable. A child can be self-directed and still have friends, mentors, projects, and structure — just not the structure of a school day.

The critics' real point is that unschooling outcomes are unevenly distributed. Children of curious, literate, well-resourced parents with flexible work do well. Children of disengaged, illiterate, or overworked parents do not, and unschooling without parental engagement collapses into neglect with a nicer name. Astra Taylor, who was unschooled and has written about it, is honest about this asymmetry. The model is a privilege multiplier when it works and an abandonment when it doesn't. This is true of homeschooling generally, but more so for unschooling, because the absence of curriculum removes the floor that even a poor curriculum provides.

For Law 2 (Think) and Law 0 (Humility), the unschooling literature is useful even for families who will never unschool. It forces the question of what schooling is actually for, which is harder than it looks. If the answer is "the things schools claim to teach," unschoolers can usually reproduce those things at lower cost and without the side effects. If the answer is "socialization into the kind of institution most adults will spend their lives in," unschoolers are doing something quite different, and the value of that difference depends on what kind of life one expects the child to lead. The honest version of the argument is that no one knows what the next forty years will require, and a child who has learned how to follow their own curiosity, fail in private, and pick up new fields without permission may be better equipped than a child who has learned how to please a sequence of teachers.