Decide what not to rebuild

A rebuilding civilisation is not starting from ignorance. Put that foreknowledge in writing: a public register of technologies and materials deliberately not adopted, each entry stating what it is, the harm, the evidence, the date decided, and the alternative used instead.

Adopt entries by the decision rule and review the register on a schedule. The test is enforcement: a convenient opportunity to break it that was refused.

This site keeps such a register: the better path lists each harmful adoption and each fix that arrived late, on the step where the decision is made, with the evidence.

Each took decades of harm to identify, and each is trivially avoidable if you simply do not adopt it:

• Tetraethyl lead in fuel, and lead paint — neurotoxic; measurably lowered children's IQ across whole populations.
• Asbestos — mesothelioma and lung cancer decades after exposure.
• PCBs — persistent, toxic insulating oils.
• CFCs — destroyed stratospheric ozone; phased out by the Montreal Protocol of 1987.
• Persistent fluorinated chemicals — non-stick and water-repellent coatings that never break down.
• Mercury-cell chlor-alkali — persistent mercury contamination.

Settlements built around the private car produced sprawl, dependence and a permanent energy demand. Density, mixed use and walkability were solved problems abandoned for reasons that turned out not to hold.

Decide in writing: walking distance to daily needs as a planning standard, freight and rail before private vehicles, and street widths and plot sizes — see plot, street and block — that keep the settlement walkable as it grows.

Some intermediate steps genuinely can be skipped: distributed generation without a coal grid; radio and networks without a copper telephone system.

Others are illusory. You cannot skip precision to reach engines, or purity to reach chips. Physical dependency chains do not care what you know. For each proposed leapfrog, write down every physical prerequisite and check it exists. If any is missing, it is not a leapfrog; it is a plan to fail.

The first industrial revolution's central accounting error was treating air, water, soil and climate as free.

Measure what each process emits and depletes — see energy without the carbon bill — assign the cost to the producer, and publish the accounts. A process that looks cheap only because someone downstream or in the future pays is not cheap. Building the true cost in from the beginning is far cheaper than retrofitting it two centuries later.

A society that can still smith, farm, build and heal at tier 2 can survive losing tier 6. Redundancy across tiers is the only real insurance, and it is cheap.

• Named practitioners and apprentices for the core tier 0–2 skills — see the record.
• Hand tools and simple equipment kept in working stock.
• The yearly lower-tier drill — see institutions that outlive people.

Letting the low tiers die out, so any step backwards becomes a fall, is a named failure.

Every historical civilisation optimised for growth until something stopped it. Whether a civilisation can choose a level and hold it is genuinely untested — and it is the most interesting open question on this site.

Write explicit targets — energy use per person, material throughput, land under cultivation, population — with the reasoning, and review them each decade by the decision rule. A target that is never written down cannot be held.

The next collapse will produce another group at tier 0. What you leave them is not just the technology — it is the reasoning, including the mistakes.

For every major choice, archive what was decided, the alternatives, the evidence, the dissent, and what happened afterwards. Leaving no record of the reasoning, so the next group inherits the conclusions without the arguments, is a named failure. Keep it in the redundant archive, in plain language.