Childhood Is an Invention

Philippe Ariès published Centuries of Childhood in 1960 with a startling claim: that medieval Europe had no concept of childhood as a distinct stage of life. Children appeared in art as small adults with adult proportions, wore adult dress, worked as soon as they were physically able, and were not marked off as a separate category requiring separate treatment. Childhood, on his account, was invented from roughly the seventeenth century onwards.

The strong version has been substantially demolished, and it is worth being clear about how. Letters, diaries, wills, coroners' records and church documents contain a great deal of unmistakable parental grief centuries before his supposed invention: parents who kept a dead child's clothes, who could not speak of it years afterwards, who made provision in wills with obvious tenderness. And painting children with adult proportions is a stylistic convention of a period that also painted adults without perspective; reading it as evidence about feeling was an error of method.

What survives is the weaker and considerably more interesting claim. The institutions built around children are genuinely modern, even though the attachment is not. Compulsory schooling, a legal age of consent, child labour law, paediatrics as a specialism, a commercial market in toys, a separate literature: every one of these is recent, and most arrived within a century.

Compulsory schooling and the factory acts did the heaviest lifting. Within two generations a child went from a contributor to household income at seven or eight to a dependant until fourteen and then later still. That is an enormous change in what a family is for economically, and it happened fast enough that both arrangements coexisted within living memory in most industrial countries.

The distinction this leaves behind is useful well beyond the topic: between an experience, which may be roughly constant across human history, and an institution, which is specific and datable. Loving a child is not obviously a historical variable. The legal, educational and medical apparatus surrounding children is, entirely. Keeping the two apart is what allows you to say that childhood was invented without saying something absurd about parents.

High infant mortality fits into that frame rather than undermining it. It did not produce indifference: the documentary evidence against that is overwhelming and comes from every period and class that left records. What it changed was planning: how many births a family expected to raise, what could be invested in, and in some places when a child was named. That is a different claim from the one about feeling, and it is the one the evidence supports.

Which brings the argument up to the present, where it is still running. Two countries set the age of criminal responsibility ten years apart. Developmental evidence constrains the plausible range and cannot pick a number, because no capacity switches on at a birthday. What each society is actually deciding is how much it is prepared to excuse, and it is expressing a moral choice in the vocabulary of development. Noticing that does not make the decision arbitrary. It does make it a decision.