Source Evaluation

Collector research rarely suffers from a total absence of information. The harder problem is deciding which records deserve weight, which claims are independent and what each source is actually capable of proving.

Source evaluation treats every catalogue, archive, database, expert opinion, market description, community post and first-hand account as evidence with a particular relationship, purpose and limit. The aim is not to find a perfect source, but to build conclusions from sources used for the right questions and weighted proportionately.

Explore source evaluation

Eight detailed topics grouped by the stage of evaluation.

8 detailed topics

Understand the source

Identify the source relationship, relevant expertise, intended purpose and incentives before relying on its claims.

Weigh supporting evidence

Judge how directly the evidence bears on the proposition and whether several records genuinely strengthen one another.

Resolve change and uncertainty

Work through disagreement, revision and missing evidence without forcing a stronger conclusion than the record supports.

The central evaluation problem

A claim can look strong because it appears in many places, carries an expert name or comes from a contemporary document. None of those qualities alone establishes that the source is accurate, independent or relevant to the exact proposition under investigation.

Strong evaluation begins by defining the claim and then asking what relationship each source has to it. A source may be excellent evidence for how an object was marketed, weak evidence for who manufactured it and irrelevant to whether the surviving example retains its original components.

A repeatable source-evaluation sequence

1. Define the proposition

State exactly what the source is being used to support: identity, date, maker, configuration, provenance, rarity, value or another bounded claim.

2. Identify the source relationship

Ask how close the source is to the object, event or transaction and whether it records, interprets, summarises or merely repeats information.

3. Match authority to the claim

Determine whether the source has relevant expertise, access and method for this precise question rather than relying on reputation alone.

4. Examine purpose and incentives

Consider why the source was created, who benefited, what audience it addressed and which facts it may have simplified, emphasised or omitted.

5. Assign evidential weight

Judge relevance, proximity, specificity, verifiability, completeness and consistency. Weight belongs to the claim, not permanently to the source.

6. Trace independence

Follow citations and wording backwards to discover whether several apparent sources represent separate evidence or one claim copied repeatedly.

7. Test conflict, currency and silence

Separate genuine contradiction from different scope, chronology or terminology, and ask whether missing evidence should reasonably have existed or survived.

8. Record a bounded conclusion

Explain what the source establishes, what it merely suggests, what remains unresolved and what further evidence could change the assessment.

Important distinctions

Primary does not mean infallible

A contemporary document or first-hand account may be close to an event while still containing errors, provisional information, bias or incomplete knowledge.

Authority is claim-specific

A leading market specialist may be strong on current demand and weak on materials analysis, legal title or manufacturing history.

Bias does not make a source useless

Purpose and incentive change how evidence should be read. A sales catalogue may be weak proof of attribution but strong evidence of how an object was marketed at a particular time.

Repetition is not independence

Many listings, databases or articles can still represent one unsupported statement when each source ultimately derives from the same original claim.

Conflict is not always contradiction

Sources may describe different examples, periods, regions, definitions or states of knowledge. Their scope must be aligned before choosing between them.

Absence is only meaningful when evidence was expected

A missing reference becomes stronger negative evidence only when the relevant source was likely to record the feature and the search was sufficiently complete.

Detailed Topics

Related Topics