History

The documentary face of the Field Guide. Answers where did all this come from, and who found it?

Parent: ../field-guide.md

What is here

The narrative is not written yet and that is deliberate - see below. timeline.md is the container it will be written from, and it exists now so that findings have somewhere to land.

Why the container exists before the narrative

This project already made this argument once. DECISIONS.md 2026-08-21, on the site register: "the register is where research output lands, so it has to exist first."

That principle was established for sites and not applied to history, so historical findings have been accumulating into documents organized for other purposes - references/primary-sources.md by evidentiary function, references/claim-genealogy.md by claim. Both are good at their jobs and neither is a place where a new date goes. The chronology had to be reassembled from scratch each time anyone needed it.

A new historical finding now has an address. That is the whole point of this file existing before anything is written.

Why the narrative waits

The leading claims are still moving. In a single day this project moved PJ-0011 889 m, relocated PJ-0016 1,082 m in nearly the opposite direction, surfaced a 1936 attestation that shifts the Symons/Shipley conflict, and raised H7 - a possible second locality on the Coyote side that no modern account mentions.

A narrative written on top of that would need rewriting, not editing.

Three cheap desk tests should land first, each of which changes what the history says rather than what it cites:

1. Locate the A. G. Swanson ranch - decides whether H7's Coyote-side locality is a chapter or a footnote.
2. Georeference Dibblee plate 4 against the corrected PJ-0016 - data/geology.md: "the whole source-geology case for PJ-0016 rests on it… the answer is genuinely unknown."
3. Run lot 16's boundary courses from the Llagas Avenue monument - settles the Kelley prospect survey-grade and the MRDS duplicate-record problem.

Trigger: write the narrative when those three resolve, or when the four interlibrary loans arrive (due before 2026-11-30) - whichever comes first. That is a condition, not a sentiment.

Why the timeline is safe to build now, when the narrative is not

timeline.md records what a source says and when, tiered. It does not record what this project concludes.

"1936: Rocks and Minerals p.46 reports the outcrop found in the creek bank about half a mile upstream of the float (tier 2)" stays true no matter how many times the coordinate moves. That discipline is what makes the container stable while the model is not - and it is the same reason the register separates a source's wording from the project's interpretation of it.

The rule that governs anything written here

Evidence-first, never narrative-first. The existing corpus on this subject collapses to roughly 13 originals endlessly repeated (references/claim-genealogy.md), and four CDMG volumes across 28 years rest on one 1934 article nobody has read. A polished narrative without its apparatus would make this project the fourteenth original, and someone would cite it as independent corroboration. That is a worse outcome than publishing nothing, because it would be credible.

Full brief for the eventual narrative: ../../prompts/history.md.