Of the Earth
The rules the schema enforces

The Covenant

An archive of land stories is an ethical problem before it is a design problem. This document exists so that the archive does not become the thing it is about.

Of the Earth is built on the Field Notes spine: one folio per encounter, plotted on a real map. Field Notes records the singer. This one records the ground that was sung to.

The salvage ethnographers of the nineteenth century collected sacred knowledge before its keepers were said to be going, locked it in institutions, and made it retrievable by everyone except the people it came from. A searchable, chunked, machine-queryable corpus of Indigenous cosmology is that same machine with better typography. Six rules stand between this archive and that outcome. They are enforced by the build, which refuses to publish a folio that breaks one.

I

The page has no single title

Every place carries every name it has been given, each with its language, the people who speak it, and what it means. The settler or survey name is one row in that list, never the headline. Where a place has a living name in the language of the people who have held it longest, that name is the record title.

names[] title

II

Refusal is content, not absence

Some stories are seasonal and told only in winter. Some are clan-held. Some are closed to outsiders by the explicit, published request of the people who hold them. The archive records that a story exists, who holds it, and that it is not ours to tell. It does not fill the hole with a collector's transcription.

  • openpublished by its keepers for general audiences
  • attributedpublic, but carried with its source community named
  • restrictedspecific detail is closed by published request, and we honour it
  • withheldwe hold a record of the place and decline to publish the story

A withheld folio renders as a visible page with the story block empty and the steward's contact in its place. The gap is the point. An archive that shows where it stops is more honest than one that pretends to be complete. The build refuses any folio marked withheld that still carries story prose.

permission withheld_note

III

Cite the living before the collector

Where a community, nation, cultural office, or contemporary scholar from that people has published on their own ground, that is the source. A colonial collector's transcription is a fallback and is labelled collector on the page, in red, so the reader knows whose hand they are reading through.

sources[].kind

IV

Every folio names who to ask

No entry is anonymous ground. Each one names the people who hold it and, where a public office exists, how to reach them. This turns a reader into a correspondent instead of a consumer.

steward.holder steward.ask

V

What is, and what can be

Every folio carries both. What is states the present condition as fact and never softens it. What can be names the real, ongoing, citable work: the dam removal, the court ruling, the seed bank, the burn that was illegal for a century and is now policy. No folio is allowed to end in despair it has not earned, and none is allowed to end in hope it cannot cite. Where there is no horizon, as at Pikinni, the folio says so plainly.

condition horizon

VI

The map draws no borders

The field map carries coastline, relief and water. It does not carry national boundaries. Not as decoration, as method: the ground in this archive predates every line drawn across it, and several of these places are cut in half by one. Basemap labels stay off.

·

On the source

The frame comes from Robin Wall Kimmerer's Braiding Sweetgrass, and from Gary Nabhan, quoted there: there is no restoration without re-story-ation. Kimmerer is Potawatomi and a botanist, and she published that thought for general readers herself. Under Rule III that makes her a community source, and the epigraph carries her name and her nation wherever it appears.

The corpus is not an attempt to hold her knowledge. It is an attempt to hold the question she asked, across as much of the world as can be sourced honestly.

·

On errors

Factual errors are corrected silently and immediately. No editor's notes, no changelog of shame. If a keeper community tells us an entry is wrong or should not be published, the entry changes or it comes down, and their reading wins over the citation. That is the whole covenant in one sentence.