BEFORE TILL archival sealBefore TillThe Living Evidence Archive

The research standard

How We Know What We Know

This archive is built to be checked. The rules below govern what may be published, at what confidence, and under what conditions a conclusion changes.

Research & Interpretation byAmberly R. Carter-Mills— Founder, Creator & Descendant Archivist

How BEFORE TILL works

From family memory to a published conclusion

  1. 01

    Family memory

    An account carried in the family enters the archive as its own category of evidence — recorded faithfully, never upgraded silently.

  2. 02

    Question

    The account becomes a researchable question with a stated answer condition.

  3. 03

    Source

    A record is located, described, and cited: repository, creator, date, locator.

  4. 04

    Evidence

    Discrete observations are extracted from the source, one statement at a time.

  5. 05

    Claim

    Observations are attached to a specific claim about a specific person.

  6. 06

    Analysis

    Supporting and conflicting evidence is weighed in writing, in public.

  7. 07

    Current conclusion

    A versioned conclusion is published with a status, a confidence, and what remains unknown.

  8. 08

    Revision

    New evidence reopens the conclusion. The change is logged with its date and reason.

Evidence labels

What each status means

Documented

Direct evidence strongly establishes the claim.

Corroborated

Multiple pieces of evidence collectively support the conclusion.

Family Tradition

Preserved within family history but not yet independently established.

Under Investigation

Evidence exists, but the conclusion remains unresolved.

Corrected

Earlier understanding changed after stronger evidence emerged.

Why some questions remain open

The archive does not fill silences

A responsible archive does not fill every silence. When surviving records cannot answer a question, BEFORE TILL leaves the uncertainty visible and continues searching.

View open questions

Principles

Six commitments

No claim without a status

Every historical statement in this archive is tagged with one of five evidence statuses before it is published. An untagged statement is a defect, not a shortcut.

Oral tradition is evidence, not proof

Family accounts are recorded faithfully and preserved as Family Tradition. They are never silently upgraded to documented fact, and they are never discarded because a record has not yet been found.

Absence is documented

Where enslavement, courthouse loss, or administrative indifference destroyed the record, the archive names the specific gap and the specific cause rather than presenting a smooth narrative.

No inference presented as fact

Dates are not estimated into birth years, relationships are not assumed from surname or proximity, and no portrait is ever substituted with a lookalike. Where a real photograph exists but survives only as a degraded reproduction, the archive may publish a digital restoration of it. Such an image is always labelled a derived likeness, is captioned with the original source it was restored from, is never presented as an archival image, and never counts toward a verified portrait.

Conclusions may change

New evidence can revise or overturn a published conclusion. Revisions are logged in the Research Ledger with a date and a reason; nothing is edited away quietly.

The family is the subject

This archive documents the Smith–Tolliver generations as people with property, labor, faith, and movement of their own — not as prologue to a single event.

Sources

Hierarchy of evidence

  1. 01Original records created at the time of the event by a participant or official witness
  2. 02Original records created near the time of the event by an informed party
  3. 03Derivative records: transcriptions, indexes, and compiled abstracts
  4. 04Published genealogies and secondary compilations, used only as finding aids
  5. 05Oral tradition and family memory, recorded and preserved as its own category

Definitions

The five statuses in full

Documented

Direct documentary evidence supports the claim.

Corroborated

Multiple independent pieces of evidence support the conclusion.

Family Tradition

The story has been preserved through family memory, oral history, or later family accounts but has not yet been independently established.

Under Investigation

Evidence is incomplete, conflicting, or actively being researched.

Corrected

Earlier family or public accounts conflict with stronger evidence and have been revised.

After publication

How new research reaches the public

  1. Discover
  2. Enter source
  3. Extract evidence
  4. Review
  5. Conclude
  6. Publish update

Drafts are never public. A change reaches a visitor only after review and publication, and it appears in the research updates with its reason. Archive version: Public Beta · Version 0.1.

Names and identity

One person, many spellings

How identity is decided

Identity is weighed on time, place, family, property and estate, associates, role, chronology, and whether another person of that name plausibly lived in the same place at the same time. Absence of an explicit statement is not absence of evidence — but it is never treated as certainty either.

Documented identity
A record explicitly connects the different name forms to the same individual.
Correlated identity
Several independent pieces of evidence strongly indicate the names are one individual, though no single record states the equivalence.
Probable identity
The evidence favours one person, but important gaps or ambiguities remain.
Possible identity
Meaningful similarities, but not yet enough evidence to merge the identities.
Separate identity
Evidence establishes that the names belong to different individuals.

The three layers

Evidence, context and reasoning

What each layer is

Context explains why a record reads the way it does. It is never evidence about this family, and it can never change the status of a claim.

Evidence, context and reasoning are three different things

Evidence is what a record actually says: a name on a line, an age in a column, a price in a ledger. It is cited to the page.

Context is the world that record was made in: the laws, the trade routes, the census instructions, the county''s politics. Context explains why a record exists and why it reads as it does. It proves nothing about any individual family.

Reasoning is the step between the two: why a researcher reads a particular record a particular way, and what reading would overturn it. BEFORE TILL publishes all three, and keeps them visibly apart. Context and reasoning are marked as such on the page, and neither can change the evidence status of a claim.

General background — not cited

Widely accepted historical background, recorded here without a single citation. It explains the record. It proves nothing about this family.

BT-CTX-00588

Why the working is published

Reasoning, not proof. Each note gives the usual reading of the evidence and then the reading that would overturn it.

Why this archive publishes its reasoning

Why show the working rather than only the conclusion?

The usual reading

A conclusion without its reasoning cannot be checked, corrected or built on. Published family trees are full of confident relationships whose basis no one can now recover, and a great many of them are wrong.

Writing the reasoning down, next to the counter-reading, means a later researcher can see exactly which step to attack when new evidence arrives — and it means this archive can be corrected in public rather than quietly.

What would overturn it

Reasoning is not a substitute for evidence, and publishing it is not a licence to lean on it. A hypothesis stated well is still a hypothesis.

Nothing in this section changes a claim''s status. Only records do that.

General background — not cited

Widely accepted historical background, recorded here without a single citation. It explains the record. It proves nothing about this family.

BT-RSN-00595

Authorship

Attribution policy

Amberly R. Carter-Mills is the founder, lead researcher, and author of this archive. Attribution is permanent and non-transferable. Research briefs, claims, corrections, and notes published here carry her name and a version date, and may be cited accordingly.

Access

Documents held but not published

Not every record that supports a conclusion can be reproduced. Modern records, and any record naming a living person, are held privately: the archive publishes the type of document, the issuing authority, what it establishes, and who checked it, so another researcher can request the same record through the proper channel. No document about a living person is ever displayed.

The records and documents policy