+++
title = "All Sixteen Were Forged"
description = "A well-sourced chain can establish that a record was made without establishing the underlying biological or economic relationship. Genealogy's source hierarchy is a discipline for naming that gap."
date = 2026-08-19
draft = false

[taxonomies]
topics = ["genealogy", "historical-documents"]
kinds = ["attestation", "provenance"]

[extra]
tier = "public"
schema_type = "Article"
canonical = ""
hero = "articles/all-sixteen-were-forged/hero.webp"
hero_alt = "Descent chart beside three supporting parish-register extracts"
diagram = "diagrams/all-sixteen-were-forged.svg"
podcast = ""
mechanism = "Genealogy grades an asserted relationship against independent evidence while historical-document work separates custody, dating and authenticity, so both treat a document as evidence that an assertion was recorded before treating it as proof that the assertion is true."
apparatus = "Genealogy imposed complete citations, relationship-specific evidence classification and explicit conflict resolution; historical-document work separately records custody, dating method and authenticity."
discriminating_test = "Ask which observation outside the document's own creation chain would contradict the claimed relationship; if none exists, the document is attestation rather than independent confirmation."
false_friend = "Forgery is a different defect: a forged document fails authenticity tests, while an authentic document can still record an incomplete, mistaken or socially convenient relationship."
test = "For every material relationship, label the support as existence, assertion, control, cash-flow or independent corroboration, and do not let a lower class populate a higher-confidence graph edge."
symptoms = [
  "A KYC file is treated as proof of beneficial control",
  "A cap-table annotation is used as proof of economic entitlement",
  "Several internally consistent documents are counted as independent confirmation",
  "Custody, date and authenticity are collapsed into one confidence label",
  "A conflicting relationship record disappears from the final ownership graph",
]
sources_verified = false

[[extra.faq]]
q = "Does a signed KYC document prove who controls a company?"
a = "No. A signed KYC document proves that a person or institution recorded a beneficial-control assertion on a particular date. Control requires an independent relationship trace, such as an appointment power, voting agreement, contract, exercised instruction, or asset-control record. Genealogy reaches the same limit when a record names a parent: the record is classified and correlated with other evidence before the parent-child relationship is treated as established."

[[extra.faq]]
q = "When do several documents count as only one source of evidence?"
a = "Several documents count as one evidentiary route when they derive their relationship claim from the same filing, register, template, intermediary, or untested assertion. Agreement moves a conclusion toward confidence only when the sources are independently derived. The required check is whether an observation outside the documents' own creation chain could contradict the relationship; without that route, repetition is attestation multiplied, not corroboration."

[[extra.faq]]
q = "Why is an authentic document not enough?"
a = "Authenticity establishes that the document is what it purports to be as an object; it does not establish that every relationship stated in it is complete or true. The Rosetta Stone is undisputedly genuine and its text has independent corroboration, while the Museum of the Bible fragments show the separate problem of forged objects entering a genuine corpus. Object authenticity and relationship truth therefore require different tests."

[[extra.faq]]
q = "What happens when a record conflicts with the ownership graph?"
a = "The conflict remains attached to the relationship and is addressed explicitly. The Genealogical Proof Standard requires evidence to be classified, correlated and reconciled rather than silently dropped. A final graph may retain an unresolved edge or state why one record is more reliable for that particular claim, but it does not convert a disputed assertion into a settled control or economic-entitlement relationship."

[[extra.faq]]
q = "Can a DNA match prove a family relationship by itself?"
a = "No. A shared DNA amount produces a ranked probability across several possible relationship categories, not a single verdict. Genealogical practice triangulates the result by building documentary trees for strong matches, finding the predicted common ancestor, and checking whether independent matches converge on it. An untriangulated match is a lead; a relationship claim requires the documentary and independent evidence that the estimate alone lacks."

[[extra.sources]]
title = "\"A More Perfect Union: The Creation of the U.S. Constitution,\" National Archives"
url = "https://www.archives.gov/founding-docs/more-perfect-union"
kind = "primary"
fetched = "2026-08-13"

[[extra.sources]]
title = "\"Code of Hammurabi,\" Wikipedia"
publisher = "Wikimedia Foundation"
url = "https://en.wikipedia.org/wiki/Code_of_Hammurabi"
kind = "tertiary"
fetched = "2026-08-13"

[[extra.sources]]
title = "\"Cotton library,\" Wikipedia"
publisher = "Wikimedia Foundation"
url = "https://en.wikipedia.org/wiki/Cotton_library"
kind = "tertiary"
fetched = "2026-08-13"

[[extra.sources]]
title = "\"David Rubenstein,\" Wikipedia"
publisher = "Wikimedia Foundation"
url = "https://en.wikipedia.org/wiki/David_Rubenstein"
kind = "tertiary"
fetched = "2026-08-13"

[[extra.sources]]
title = "\"Dead Sea Scrolls,\" Wikipedia"
publisher = "Wikimedia Foundation"
url = "https://en.wikipedia.org/wiki/Dead_Sea_Scrolls"
kind = "tertiary"
fetched = "2026-08-13"

[[extra.sources]]
title = "\"Donation of Constantine,\" Wikipedia"
publisher = "Wikimedia Foundation"
url = "https://en.wikipedia.org/wiki/Donation_of_Constantine"
kind = "tertiary"
fetched = "2026-08-13"

[[extra.sources]]
title = "\"Dunlap broadside,\" Wikipedia"
publisher = "Wikimedia Foundation"
url = "https://en.wikipedia.org/wiki/Dunlap_broadside"
kind = "tertiary"
fetched = "2026-08-13"

[[extra.sources]]
title = "\"Magna Carta,\" National Archives (featured documents)"
url = "https://www.archives.gov/exhibits/featured-documents/magna-carta"
kind = "primary"
fetched = "2026-08-13"

[[extra.sources]]
title = "\"Museum of the Bible,\" Wikipedia"
publisher = "Wikimedia Foundation"
url = "https://en.wikipedia.org/wiki/Museum_of_the_Bible"
kind = "tertiary"
fetched = "2026-08-13"

[[extra.sources]]
title = "\"Rosetta Stone,\" Wikipedia"
publisher = "Wikimedia Foundation"
url = "https://en.wikipedia.org/wiki/Rosetta_Stone"
kind = "tertiary"
fetched = "2026-08-13"

[[extra.sources]]
title = "\"Vinland Map,\" Wikipedia"
publisher = "Wikimedia Foundation"
url = "https://en.wikipedia.org/wiki/Vinland_map"
kind = "tertiary"
fetched = "2026-08-13"

[[extra.sources]]
title = "Board for Certification of Genealogists, \"Genealogical Proof Standard\""
url = "https://bcgcertification.org/ethics-standards/"
kind = "secondary"
fetched = "2026-08-13"

[[extra.sources]]
title = "DNA Painter, \"Shared cM Project v4\""
url = "https://dnapainter.com/tools/sharedcmv4"
kind = "secondary"
fetched = "2026-08-13"

[[extra.sources]]
title = "Magna Carta 1297, c.9, legislation.gov.uk"
url = "https://www.legislation.gov.uk/aep/Edw1cc1929/25/9"
kind = "primary"
fetched = "2026-08-13"

[[extra.sources]]
title = "Wikipedia, \"Board for Certification of Genealogists\""
publisher = "Wikimedia Foundation"
url = "https://en.wikipedia.org/wiki/Board_for_Certification_of_Genealogists"
kind = "tertiary"
fetched = "2026-08-13"

[[extra.sources]]
title = "Wikipedia, \"Elizabeth Shown Mills\""
publisher = "Wikimedia Foundation"
url = "https://en.wikipedia.org/wiki/Elizabeth_Shown_Mills"
kind = "tertiary"
fetched = "2026-08-13"

[[extra.sources]]
title = "Wikipedia, \"Forensic genealogy\" (investigative genetic genealogy)"
publisher = "Wikimedia Foundation"
url = "https://en.wikipedia.org/wiki/Forensic_genealogy"
kind = "tertiary"
fetched = "2026-08-13"

[[extra.sources]]
title = "Wikipedia, \"Genealogy\""
publisher = "Wikimedia Foundation"
url = "https://en.wikipedia.org/wiki/Genealogy"
kind = "tertiary"
fetched = "2026-08-13"

[[extra.sources]]
title = "Wikipedia, \"Genetic genealogy\""
publisher = "Wikimedia Foundation"
url = "https://en.wikipedia.org/wiki/Genetic_genealogy"
kind = "tertiary"
fetched = "2026-08-13"
+++

Around 2002, roughly 70 fragments resembling Dead Sea Scrolls material appeared on the antiquities market from undisclosed sources, with no connection to anything recovered at Qumran between 1946 and 1956. The Museum of the Bible acquired 16 of them and put them on display as genuine.

Germany's Federal Institute for Materials Research and Testing found five to be modern forgeries in 2018. By March 2020 all 16 had been confirmed forged.

That finding is decisive about those sixteen objects and about nothing else.

It does not follow that a genuine fragment recovered at Qumran proves whatever authorship, transaction or relationship a reader wants to put behind its text. The inverse matters more, and gets far less attention: an authentic object can carry an assertion that was selected for convenience, transmitted at a remove, or simply incomplete for the relationship being investigated.

A document can be genuine, correctly dated and continuously held, and still be the wrong document for the claim being made from it.

## Genealogy asks the question first

That distinction is ordinary in historical-document work. It is more demanding in genealogy, where the object of the claim is a relationship nobody can put back on the table.

So the discipline begins by narrowing the question until it can be answered at all. The identity of a child's father. The birthplace of an immigrant. Whether two same-named men in one county were one man.

None of those is the question anybody arrived with. Each is a question a record can bear on.

Elizabeth Shown Mills' *Evidence Explained* supplies the administrative half, and the administrative half is where the transfer lives. A citation recorded at the point of use carries the record's status and its access point along with the claim.

That means a second researcher can go back to the same source, read what it actually said, and test whether it supports the relationship or merely repeats it. Without the citation the assertion travels and the evidence does not.

None of this is a hierarchy of better and worse paper. It exists to block one specific conversion: the fact that an official-looking record exists becoming the conclusion that the relationship it recites exists.

## Six records, one assertion

The tempting answer to a weak relationship is more documents. Genealogy asks first whether the new document is independent of the old ones.

Agreement across independently derived records raises confidence. Agreement among copies, extracts, later compilations and records built from the same original declaration does not become independence by accumulation, however many containers it arrives in.

The form of the conclusion is where that becomes visible. A clean case with no conflict gets a proof statement. A moderately complex one gets a proof summary. Conflicting records, indirect evidence or a risk of identity collision require a proof argument — a connected account naming the evidence, setting out each conflict, and explaining the resolution. The length of the write-up is itself evidence about the size of the uncertainty.

DNA does not waive any of this. The Shared cM Project reports a shared total as a distribution over possible relationships rather than a single answer. Genealogists build documentary trees for strong matches, look for the common ancestor the estimate predicts, and require convergence from independent matches. A shared segment is a lead until that work is done.

The transfer is exact. A cap-table note establishes that somebody recorded an entitlement. A KYC file establishes that a customer made a declaration about beneficial control. Neither creates the cash flow, the voting power, the appointment right or the exercised control that would confirm the economic relationship independently.

Six systems that have each copied the same declaration are one route back to one assertion.

## Different questions, different fields

Historical-document work keeps the questions in separate columns because none of them substitutes for another. Medium, dimensions, language, custody, date and dating method are all recorded, and all recorded apart.

The British Museum's entry for the Rosetta Stone describes a granodiorite slab of stated dimensions and weight. Those measurements identify the object. They say nothing about the political relationships named in its decree.

The date is a further claim again, and the method is part of it. Hammurabi's code is dated by content and archaeological context. The Dead Sea Scrolls are dated by paleography and radiocarbon, corroborated by coins found at Qumran. Naming the method states what kind of date has been obtained, which is also a statement about what the date cannot do.

Custody is another column entirely. The Rosetta Stone has been held by the British Museum since 1802, and Egypt has requested its return since 2003. That dispute is about where an acknowledged genuine object belongs. It is not a test of whether the object is genuine, and a custody chain proves possession rather than any relationship a reader infers from the thing possessed.

Sixteen forgeries were caught because somebody put them under an instrument built to catch forgeries.

Nothing puts a genuine document under an instrument for the relationship it was asked to prove. It is authentic, it is cited, and the edge it created goes on carrying weight in three systems that never saw it.