Report 002026-05-30

Long Family Archive. Audit Report (2026)

A read-only verification and relabel of the archive against FamilySearch and published genealogical scholarship. No genealogical content was deleted; claims were re-tiered, cited, and, where a claim is disputed in published scholarship, relocated to the archive basement as labeled family tradition.

Methodology principle: lead with what was found, not with what was hoped for. Every claim carries a confidence tier; where the evidence runs out, the archive says so.


What the audit checked

  1. Inventory (01-inventory.json, 01-inventory.v2.json), every named ancestor and FamilySearch PID rendered across 70 pages, plus the parent-child chain edges the site asserts.
  2. FamilySearch verification (02-fs-verification.json / .md), every unique PID resolved against the FamilySearch beta Family Tree API, with source classification per person and per parent-child edge. Read-only; no FS data modified.
  3. Gateway cross-reference (03-gateway-crossref.json / .md), every pre-1750 ancestor checked against accepted gateway scholarship (Hereditary.us / Taylor & Murphy; Richardson; Roberts; Weis; Faris; the General Society of Mayflower Descendants).

Key findings

People are well-sourced; the edges between them are not

  • 366 unique PIDs resolved. Person tiers: 318 verified, 30 sourced, 15 FS-tree-only, 3 unresolved (2 living, 1 merged).
  • 115 unique parent-child edges verified. Edge tiers: 0 verified, 0 sourced, 91 FS-Tree-Only, 24 Gap.
  • The central finding: a person can carry dozens of primary-record sources on their profile while the relationship to their parent carries none. FamilySearch records the edge as a tree assertion, not a sourced conclusion. Across the whole archive, no parent-child edge is documented at the relationship-conclusion level. This is why the site now tiers the edge separately from the person, and introduces the FS-Tree-Only tier.

The royal/Plantagenet descent is not accepted scholarship

  • The site's single royal-descent claim runs through the Rhode Island immigrant Stukely Westcott up to the Plantagenet kings (Edward IV via Arthur Plantagenet), on the unproven premise that his mother or grandmother was a Stukley of Devon.
  • Ten independent sources, 1886-2026, including the WikiTree Edward III Gateway Ancestors Project, conclude this splice is not documented. Stukely Westcott appears in none of Richardson's Royal Ancestry, Roberts' Royal Descents of 900 Immigrants, Faris, or Weis.
  • Action: the Plantagenet descent was moved off the main site into the archive basement (/basement#plantagenet-claim) as labeled, disputed family tradition with full dispute citations; the standalone royal ancestor and cousin pages were removed and redirect there; all royal framing was removed from outward-facing pages, meta, and heraldry.

The accepted gateways are the Mayflower passengers

  • The only ancestors that appear in an accepted gateway compilation are four Mayflower passengers. Francis Cooke, Richard Warren, George Soule, and Degory Priest, verified via the General Society of Mayflower Descendants. They are Mayflower-lineage gateways, carrying no proven royal descent.
  • Tristram Coffin, Peter Folger, and the other immigrant founders are documented colonial ancestors but are not accepted royal-descent gateways.

Resolved findings (2026-05-30)

Hannah Coffin (M7FZ-T6T) parentage: RESOLVED. The audit framework's deep edge-sourcing pass (2026-05-30) surfaced a conflict between two compilations on whether Hannah's father was John or James Coffin. Subsequent web research confirmed James Coffin via three independent sources: Find a Grave Memorial for Richard Gardner Jr. (cited gravestone-level), Coffinquest Rootsweb (listing Hannah among James's children with siblings Samuel and Deborah), and Geni's Mary Coffin profile (Hannah's sister, same parents). The archive's previous "through John, not James" framing has been corrected to "through James, not John." The corrected edge has been added to the sourced list in audit/04. Documented example of the audit framework catching and resolving an inherited error.

Downstream follow-up (queued as Audit Pass 8): the cousin-distance summaries that were computed on Tristram as the shared ancestor (Lucretia Coffin Mott, General John Coffin, Admiral Sir Isaac Coffin, Charles A. Coffin, and John Coffin of Quebec) now share James Coffin (one generation closer). Those distances are flagged pending recomputation; the direct-line father-fact has been corrected to James everywhere, and the distance numbers carry a visible pending note rather than invented values.

Benjamin Franklin cousin distance: RESOLVED at 1st cousin, 10 times removed. This figure was corrected twice. An interim S62 edit had stated 1st cousin, 9 times removed, applying the standard NEHGS cousin-math to an eleven-generation chain count. A FamilySearch relationship-view verification on Peter Folger (May 2026) confirmed the maternal chain is twelve generations from John back to Peter Folger, so the canonical figure is 1st cousin, 10 times removed (cousin degree = min(12, 2) minus 1 = 1; removed = absolute value of 12 minus 2 = 10). That is consistent with the Maj. Gen. Benedict Arnold V distance of 4th cousin, 7 times removed, computed on the same twelve-step depth via the shared Westcott ancestor. The S62 nine-times figure was a good-faith application of the right method to a one-generation undercount; it was reversed across every surface and the correct figure restored. The correction history, formerly carried in an "Honest Acknowledgments" section on /lines/folger, now lives here; the line page states the figure and links to this report.


Confidence tiers used across the site

TierMeaning
VerifiedEdge confirmed in FamilySearch and backed by a primary or derivative record, or held in the family archive
SourcedThe edge has some source, but none primary
FS-Tree-OnlyAsserted in the FamilySearch Family Tree (a collaborative wiki tree) with no source on the relationship conclusion
TraditionalOral history or traditional genealogy
Traditional, DisputedTraditional, with a published scholarly dispute (shown with a ⚠ and a distinct color)
Research-in-ProgressActively being investigated
Inherited EvidencePrimary-source artifact in the family archive
GapNo PID, or FamilySearch does not assert the edge

Full machine-readable data: audit/02-fs-verification.json, audit/03-gateway-crossref.json. This report and the relabel were produced in the 2026 audit pass; the verification scripts are retained alongside for re-runnability.