Observation
Describe what a system, product, law, interface, or study actually does. Do not smuggle a conclusion into a technical description.
Editorial method
This is an argumentative archive, not a neutral database and not a machine-generated authority. Its standards are visible because the work depends on readers being able to distinguish evidence from interpretation.
Describe what a system, product, law, interface, or study actually does. Do not smuggle a conclusion into a technical description.
State the inference and the value judgment. A reader should be able to disagree with an argument without having to deny the observation it begins from.
Link claims to records that readers can inspect. A source note says what a document is useful for and what it does not establish.
Do not hide the strongest reasonable objection. The welfare section exists because uncertainty deserves more than a caricature.
Working standards
When a question is unsettled, the archive says so. It distinguishes research preprints from law, a product behavior from a report of experience, and a user’s real feeling from an unproven claim about a system’s inner state. It does not use a citation to imply that every author shares the archive’s position.
When a word does too much work, the archive breaks it into parts. Agency is not accountability. Legal personhood is not moral patienthood. A prediction is not an intention. A system’s apparent personality is not the same thing as a person’s stake in a relationship.
The aim is not to make disagreement disappear. It is to make the disagreement more legible, and to keep human responsibility visible when a technical system is introduced into a human setting.
Read the works →