System states

The system defaults to the safest possible behavior when uncertain, prioritizing stopability.

The system defaults to permissiveness under failure—sometimes necessary, sometimes dangerous. Must be paired with velocity friction.

A harmful state where systems fail without signaling it; the worst possible form of failure because it hides moral latency.

Originally: machines absorb force so people survive. Digitally: people absorb system failures so machines stay smooth. Ethotechnics reverses this direction of impact.

Places in a system where harm occurs but no one can see, trace, or intervene. Closing dead zones is a goal of oversight horizons.

A scheduled calm state where teams intentionally slow or stop throughput so inspections, upgrades, and rehearsals can happen without crisis pressure. Maintenance windows make stopability routine instead of reactive.

Each window is negotiated with the people impacted, includes published service guarantees, and documents which safeguards were tested so unfinished work rolls into the shared repair log.

A facilitated reflection held while the system is still in a warning band to examine how maintenance load, emotional labor, and unresolved incidents are accumulating. Care retrospectives combine telemetry with frontline testimony.

They redistribute responsibilities before burnout or harm escalates, triggering new maintenance windows or policy fixes when the team cannot keep absorbing risk.

A living record of every mitigation, decision, and resource commitment made after a fault. Repair logs make accountability legible by linking people harmed, who intervened, and what evidence was used.

They inform future care retrospectives, power audits, and service-level reports so follow-up work is traceable and burden does not drift back to the same communities.