When is criticism a gift and when is it just fear?

The test every objection runs through

If criticism holds up from first principles, it gets built in. If it is only fear, it is taken seriously and answered precisely.

The test only works if both ways are genuinely open. first-principles here means: an idea proves itself against what is fundamentally true for children and their education, not against whether it works elsewhere. Criticism that hits that ground changes the concept. Twelve such assumptions stand in the chapter, each with a correction and an if-then rule that settles in advance what happens if the risk arrives.

I have practised this check before reacting for a long time, just on a much smaller subject. In a video series from 2016 on dealing with anger I describe the same first step: first check whether I have actually understood a situation correctly, then ask lovingly instead of accusing. The setting has changed, from my own anger to criticism of a school concept. The first step has stayed the same: understand first, then answer.

• More structure automatically produces more maturity. Correction: maturity grows through relationship, real responsibility, a kept word, repair, reflected freedom.
• Every child needs one big life project. Correction: in daily life we say important project. At the start it only has to matter enough to work on it.
• Mixed age groups automatically help everyone. Correction: younger children can be steamrolled, older ones turned into assistants, roles fixed, attention shared unevenly.
• Every conflict is a learning chance for joint clarification. Correction: only a roughly symmetrical everyday dispute suits mediation. In protection cases adults investigate, decide, document, review.

What a ceiling means

A mentor ratio of 1:12 carries financially better than a much denser one. That makes it the ceiling for what one mentor can be responsible for, and a ceiling differs from the ideal size of a learning intervention. For targeted support the evidence is clearer. Small group tuition means two to five children, diagnosed on purpose, with effectiveness dropping above six or seven. The Education Endowment Foundation keeps reducing class size as a separate question.

Telling parents that 1:12 is scientifically optimal. That would not be covered by the evidence.

That honesty has a consequence. Since 11 August 2026 the rule is: every child gets 18 plus 2 minutes of conversation every week. For reading, maths gaps and specific skill sprints we form much smaller groups. The ceiling stays what it is, and the learning happens in the size where it works. Why the child and mentor relationship is the measure at all stands in the triad of relationship, meaning and self-determination.

What we celebrate about failure

Convenience disguises itself as failure easily. So we celebrate an honest setback, not the absence of an attempt. A project may fail if someone genuinely tried, sought feedback, documented and reflected. If a child stands there empty-handed on expo day, we ask for the traces: which real attempts, which data, sketches, responses, mistakes. If there is no trace at all, we talk honestly about the missing ownership.

A second parent question in the same chapter aims at the maturity points, and the answer opens with an admission: done badly, yes. That is why only a voluntary spoken commitment counts. No rankings, no points for performance or behaviour, no deductions, no automatic unlocking. What hangs on that stands in Is belonging something you have to earn?

Ten questions before anything is introduced

The same measure applies to new ideas, including our own. The chapter ends with a test chain of ten questions. Three carry most of the weight.

If an idea is weak on three or more questions, it gets sharpened first and tested again. The last question is the most uncomfortable, because it demands a date. Someone who names no day on which they could prove themselves wrong has an opinion.

The same caution sits in the wording we use outwards. The school calls itself the first Paideia partner school and nothing more, because quality criteria for partner schools do not exist yet. A sentence that promises less lasts longer. What this looks like when a tool gets too heavy stands in What do you delete before you automate? The general version lives in the open book World Modeling.

• Does it strengthen the growth relationship?
• Does it take load off mentors, or only create new administration?
• If it fails: how do we notice after eight weeks?

On record

Paideia core handbook, chapter 16 (Red team, first-principles questions and critical parent answers) and chapter 17 (Scientific review). Decision on the mentor ratio of 11 August 2026.

• The core rule of chapter 16: criticism that holds up from first principles gets built in, fear is taken seriously and answered precisely.
• Twelve named wrong assumptions, each with a correction and an if-then rule
• 1:12 is the hard ceiling for mentor responsibility, not an ideal group size
• Since 11 August 2026: 18 plus 2 minutes of conversation per child per week
• What gets celebrated is an honest setback after a genuine attempt
• A test chain of ten questions decides whether an idea is introduced or sharpened first

Criticism guide

The questions before giving and after receiving criticism, on one card. PDF · A4, 1 pages, 32 KB, as of 3 September 2026.

Sources and links

This note keeps growing

2026-09-03: Deepened with an episode of my own emotions compass series from 2016: the same first step, understand first, then answer, on a much smaller subject.

2026-09-03: Grown: the full test rule, four of the twelve named wrong assumptions, the small group evidence with the limit the handbook puts on it, the test chain and the criticism compass as a tool.

2026-09-02: Planted from the chapter on red team and first-principles questions.