Teach the next generation

Take children out of production for part of every day. This feels unaffordable and is the highest-return investment available. Three to four hours a day for children from about 6 to 14, at a fixed time, protected by the decision rule so it is not cancelled whenever work is short.

At peak harvest, shorten the day rather than stop it. A settlement that works its children full-time is choosing a lower ceiling.

One teacher to roughly fifteen children. Teaching rotates through the literate adults — each teaching a subject they actually practise — with one person responsible for the school as a whole.

Train teachers: a new teacher watches, then co-teaches, then teaches while watched. Every critical trade-holder teaches for part of every week, which is also how the settlement discovers who explains well.

Without exception — not only the children of whoever holds status. Numeracy in particular gates surveying, construction, chemistry, medicine and every kind of accounting. See arithmetic for everyone.

Set milestones by age and check every child against them: reading a simple passage by about 8, reading an unseen practical text and doing the four operations by about 10, fractions, proportion and measurement by about 12. A child behind a milestone gets more time, not less.

On top of school, each critical trade takes learners on a known term — several years for a complex trade — with written obligations both ways:

• The master teaches, on a stated schedule, and does not use the apprentice only as labour.
• The apprentice works and learns, and at the end of the term teaches the next learner.

Apprenticeship as unpaid labour with no actual instruction is a named failure: check it by asking the apprentice to demonstrate, not by asking the master.

Teach from the library, not only from memory. The point of holding books is that they let you teach things nobody present has ever done.

Write local textbooks too: this valley's frost dates, this clay's shrinkage and firing temperature, this ore's yield, this spring's flow, this group's history. That knowledge exists nowhere else and is worth more to your descendants than a general text. Copy them — see writing.

A society that has the scientific method can regenerate lost recipes; one that has only recipes cannot. Build it into lessons as practice, not as a slogan:

• Change one thing at a time, and keep everything else the same.
• Write down the conditions — quantities, times, temperatures — before looking at the result.
• Compare against a control — the same thing without the change.
• Record failures as carefully as successes.

Give older children real questions to test — which bed yields more, which clay cracks less. See controlled comparison.

Include how the group got here — what was tried, what failed, who died of what and why, which decisions were wrong. Groups that mythologise their founding lose the actual lessons, and a child who learns that the founders never made mistakes learns nothing about avoiding them.

Teach it from the written record — see the record — not from recollection, which smooths the failures away.

The school is proven when someone born in the settlement learns a skill from a book that no living member has practised — and demonstrates it.

Set this deliberately as a goal for older students: pick a documented skill nobody holds, give them the books and time, and evaluate the result against the book's own tests. It is the one test that shows the ladder will keep being climbed after the founders are gone.