Childhood, education and the transmission of knowledge

An old, untidy classroom, full of children reading, playing or squabbling around a teacher. — “A School for Boys and Girls”, painting by Jan Steen (1670). Jan Steen, Public domain, source

School is neither the only way of passing knowledge from one generation to the next, nor the oldest.

Learning by doing

In Djenné, in Mali, the great mosque is built of earth. Every year, the whole town comes to replaster it: the render is prepared, people climb the poles that jut out from the walls, and they fill in what the rain has washed away.¹

This is not just maintenance. This is how the craft is passed on. The young climb up with those who know, and learn by practising, a little more each year.²

Nothing is written down, and no one sits an exam. Yet the know-how has been handed down through the generations for a long time, because it is used.

The same principle, with a contract

In Europe, in the Middle Ages and afterwards, people also learned a trade by practising it. But a written agreement was signed.

A child would go to live with a master for several years. The child’s parents signed a contract with this master, and the trade’s guild oversaw the agreement.³

This contract set out the obligations of each party. The master lodged, fed and clothed the child and taught them his trade. Some contracts also required him to send the child to learn to read and write, or to provide the child’s tools. The apprentice, for their part, had to obey and abide by rules of conduct.

Girls were apprentices too, for instance in some textile trades.⁴

A caveat is needed here, and it applies to all documents. A contract says what was supposed to happen. It does not say what actually happened.

Apprenticeship is often portrayed in two opposite ways: as an opportunity offered to a poor child, or as the exploitation of a defenceless child. The documents bear out neither. There were real obligations on the master’s side, and real constraints on the child’s side. The rest depended on the people involved.

Learning by heart — and it is a technique

It is sometimes said that learning by heart means learning mindlessly. One case shows the opposite.

In South Asia, certain very long texts have been passed down orally over the millennia. To keep them from becoming distorted, people did not merely make pupils repeat them. They invented several interlocking ways of reciting them.⁵

First, the text straight through. Then word by word, each word taken on its own. Then in pairs. Then in braided forms, where the words come back in crossed orders, forwards and backwards.

The method is ingenious. If someone forgets a word, it no longer matches the other ways of reciting. The error reveals itself through simple comparison.⁶

It is a way of checking without writing anything down — something that writing, on its own, does not allow: a tired copyist copies a mistake without anything alerting him to it.

School — older and younger than we think

Two surprises, once again pointing in opposite directions.

Older than we think. The oldest educational institution still in operation is neither in Oxford nor in Bologna. It is in Fez, in Morocco: al-Qarawiyyin, founded in the ninth century. It predates the European universities by several centuries.

Moroccan tradition attributes its founding to a woman, Fatima al-Fihri, who is said to have devoted the money she had inherited to it.

How do we know? Almost everything we know about her comes from a single chronicle, written several centuries after the events. This is the case for many foundations of that period, whether they are attributed to women or to men.

Younger than we think. Compulsory schooling, the kind the state requires people to attend, is recent: most major countries introduced it only in the nineteenth century, some in the twentieth. France, then, is not a forerunner. It follows.⁷

The motive is surprising. Compulsory schooling did not spread primarily because children had been recognised as having a right to knowledge. Military defeats had a great deal to do with it: several major school reforms come after a lost war, and aim to bind the country together as much as to educate it.⁸

This does not mean that school served only that purpose. These laws taught millions of people to read, and they took children out of the workshops. The motive of those who vote for a law and the effect it produces are two different things, and neither cancels out the other.

One last remark, and it matters. Japan, in 1872, builds its school system by drawing on several Western models, the French one among them. Then it changes course several times. This is not a school system passively received: it is a country that chooses, tries, and changes.⁹

Meanwhile, children were working

In France, the first law limiting child labour dates from 22 March 1841.

It sets the minimum age at eight. From eight to twelve, eight hours a day at most. From twelve to sixteen, twelve hours.

One article of this law is surprising today: children aged eight to twelve had to attend school in order to be allowed to work. School was not the opposite of work. It was its precondition.¹⁰

This law applied only to mechanised factories and to workshops with more than twenty workers. Above all, enforcing it was entrusted to local notables, serving as volunteers and appointed by the prefect — that is, to men from the same society as the employers they were supposed to inspect.¹¹

They hardly did so. In 1847, a politician, Baron Dupin, pointed out to his colleagues that the law was failing. It was not until 1874 that real labour inspectors, paid by the state, came into being.¹²

A lesson worth applying often: a law is not a fact. It is a written intention. Faced with a law, the question to ask is always: who was supposed to enforce it?

It would be wrong to conclude that it achieved nothing. It was the first time the French state had decided to intervene between an employer and the workers. What a text opens up counts as much as what it fails to achieve.

What the archives do not say

Before going further, it is important to know the main flaw of this history: the archives are written by adults.

We have apprenticeship contracts by the thousand, laws, registers, school regulations. All of them say what adults expected of children. Almost nothing says what children did and thought. Children write little, and what they write is not kept.

This does not prevent us from doing history. It obliges us to know what we are talking about: what we know is mostly the rules, not the lives.

A famous thesis, and a wrong one

In 1960, the French historian Philippe Ariès published a book in which he argued that in the Middle Ages the “sense of childhood” did not exist: the child was supposedly treated as a miniature adult. The book’s success spread far beyond the circle of historians.¹³

Since then, specialists in the Middle Ages have shown that this was wrong. They have found traces of affection for young children, a real concern to educate them, and even a particular way of burying them. Ariès, moreover, was not a medievalist.¹⁴

Why does his thesis still circulate? The answer is instructive: the thesis had left the circle of historians very early on. Sociologists, psychologists and a wide public had adopted it. The criticisms, for their part, appeared mainly in historians’ journals.¹⁵

A correction does not travel as far as the error it corrects.

It should be added that Ariès listened to his critics. When he reissued his book in 1973, he qualified several claims in his preface — while maintaining the core of his thesis.¹⁶

One last clarification, because this is exactly the kind of error to be avoided: Ariès was not wrong about everything. His book showed that our way of seeing childhood is in no way self-evident and that it has changed over time. That remains true. It is the specific thesis he drew from it that does not hold up.

Today

Children still work. They are counted.

According to the joint estimates of the International Labour Organization and UNICEF, published in June 2025, nearly 138 million children aged 5 to 17 were working in 2024 — 59 million girls and 78 million boys, or almost eight per cent of all the world’s children. Among them, 54 million do work that is hazardous to their health or safety: two in five.¹⁷

They numbered 246 million in 2000. The number has thus almost halved. But states had set themselves the goal of ending it by 2025, and they have not succeeded: at the pace of the last four years, progress would have to be eleven times faster to get there in five years.¹⁸

Child labour is concentrated in agriculture — 61% of cases — followed by services and industry.¹⁹

What this figure does not count: It rests on an economic definition of work. It therefore leaves out a large part of the tasks done at home — housework, fetching water, looking after younger children. Yet these tasks fall mainly on girls.
The figure is accurate. It is also incomplete. And what it leaves out is not distributed at random.²⁰

Note 1

Le bien est caractérisé par un usage intensif et remarquable de la terre particulièrement dans son architecture. La mosquée exceptionnelle, de grande valeur monumentale et religieuse, en est un exemple éloquent.

The property is characterised by an intensive and remarkable use of earth, particularly in its architecture. The exceptional mosque, of great monumental and religious value, is an eloquent example of this.

UNESCO, “Villes anciennes de Djenné”, Liste du patrimoine mondial

Every year the structure is replastered with mud in a celebration that brings out all of the city's residents.

“How the City of Mud Stays Standing: Meet the Masons of Djenné, Mali”, Smithsonian Magazine, 30 September 2013