Conflicting Sources

A source conflict exists when two or more evidence-based claims cannot all be accepted in their present form. In collecting, the disagreement may concern maker, date, issue, completeness, authenticity, provenance, sale status, rarity or the significance of a physical feature.

The collector's task is not to make every source agree. It is to identify exactly what is disputed, establish where each claim came from, test what each source was capable of knowing, compare the claims with the object and other evidence, and record a conclusion whose confidence matches the evidence.

Conflicts are therefore not merely obstacles. They often expose hidden assumptions, different definitions, copied traditions, composite objects, transitional production or gaps in the research question itself. A careful disagreement can produce a more accurate catalogue record than an apparently tidy answer.

Core collector principle

Resolve the claim, not the discomfort

A trustworthy record does not hide disagreement to appear decisive. It shows what was claimed, what was observed, where each assertion originated, which evidence carries the greatest weight, why the current interpretation is preferred and what remains unresolved.

The boundary between fact, source claim, interpretation and uncertainty should remain visible throughout the research record.

Orientation

The four layers of a conflict record

Confusion grows when observation, quotation and judgement are blended into one sentence. Separate these layers before comparing sources.

Observable fact

What can be directly seen or measured

Marks, codes, dimensions, materials, construction, component relationships, wear and documented transaction details belong here before interpretation is added.

Source claim

What a particular source says

Record the claim with its author, issuing body, date, edition, page, lot, archive reference or saved record. A claim is not automatically a fact because it appears in a catalogue or database.

Interpretation

What the evidence may mean

An interpretation connects observations and source claims: for example, a code may indicate a production date, artwork revision or inventory sequence. The reasoning should remain visible.

Confidence

How strongly the conclusion is supported

Confidence should be stated separately from the conclusion. A specific interpretation can still be tentative, while a broad conclusion can be strongly supported.

Recognition

Not every difference is a contradiction

Before weighing evidence, test whether a real conflict exists. Different questions, levels of description, markets, production stages, components, approximations and naming systems can create apparent disagreement.

Contradiction

Claims cannot both stand as written

One source identifies a 1980 first UK release; another identifies a 1982 first UK release. The disputed event and territory are the same, so the claims require testing.

Omission

One source is silent

A catalogue that does not mention an insert does not automatically contradict a checklist that includes it. Silence becomes meaningful only where inclusion would reasonably be expected.

Uncertainty

The sources are cautious rather than opposed

Two sources that both say 'probably 1980' may share uncertainty without making conflicting claims.

Variation

Different examples may both be genuine

Different component combinations, marks or packaging can indicate regional, transitional or production variants rather than error.

Revision

Knowledge changed over time

An older attribution and a later corrected attribution may represent a documented change in scholarship, not two equally current conclusions.

Conflation

The records may describe different objects

Generic titles, reused catalogue numbers, missing photographs and grouped lots can cause two separate objects or issues to be treated as one.

Myth versus reality

More sources do not automatically make the majority right

Myth

Ten websites outweigh one archive record because ten is a larger number.

Reality

If the websites copied one guide or auction listing, they form one dependent source chain. Evidence is weighted by relevance, independence and quality, not counted by appearance.

Diagnosis

Common forms of source conflict

Direct factual conflict

Different answers to the same defined claim

Dates, dimensions, production locations, edition labels or ownership statements directly diverge. This is the clearest conflict, but it may still be resolved by checking definitions and object identity.

Definition conflict

The same words carry different meanings

Terms such as first edition, issue, state, complete, original or prototype vary across fields, territories and reference traditions. Define the disputed term before weighing the sources.

Object-document conflict

The surviving object challenges the written record

A later component date, unexpected manufacturing process or inconsistent mark may contradict a guide or seller description. Object evidence can be strong, but replacement, repair and mixed assembly must be considered.

Contemporary-source conflict

Period records disagree

Advance catalogues, advertisements, production records, registrations and retail lists may describe plans, legal dates or marketed products rather than the same event.

Later-reference conflict

Guides and databases classify differently

Authors may have seen different examples, inherited earlier mistakes, used competing numbering systems or revised their conclusions between editions.

Provenance conflict

Ownership chains do not align

Owner names, dates, lot numbers and object descriptions may be inconsistent or incomplete. Similar objects are easily conflated when records are compressed.

Market-record conflict

Sale data describes different price concepts

Hammer price, buyer's premium, tax, currency conversion, estimate, asking price and sale status must be normalised before figures are treated as contradictory.

Expert-opinion conflict

Specialists interpret the same evidence differently

Differences may arise from expertise, comparanda, classification traditions, certainty thresholds, incentives or legitimate uncertainty in an immature field.

Dependency conflict

Apparent corroboration is copied

Many websites may descend from one auction description or guide. Identical wording, errors, crops and unexplained dates reveal one source chain rather than independent support.

Composite-object conflict

Components carry separate histories

Boxes, manuals, inserts, labels and accessories may have different dates because of transition, replacement, mixed stock or later collector assembly.

Explanation

Why sources conflict

A contradiction does not itself prove negligence or fraud. Begin with ordinary explanations, then escalate the level of concern when evidence, stakes or behaviour justify it.

Ordinary error

A precise statement can begin with a simple mistake

Transcription errors, misread handwriting, swapped captions, bad conversions and incorrect cross-references can produce durable false claims.

Different access

Researchers did not see the same evidence

One source may have examined the object or archive; another may rely on a crop, summary, incomplete example or inherited description.

Evidence appeared later

An older conclusion was reasonable at the time

New records, better images or newly documented examples can overturn an earlier interpretation without implying careless research.

Terminology drift

Collector language changes

Manufacturers, retailers, early collectors, current communities and grading services may use the same label differently.

Production complexity

Manufacture rarely follows a clean sequence

Overlapping runs, reused packaging, mixed stock, multiple factories and gradual component changes can create genuinely inconsistent-looking evidence.

Purpose and incentive

Sources were created to do different jobs

Advertising, inventory, legal registration, sale promotion, donation, insurance and personal memory each prioritise different information and may omit or simplify other facts.

Institutional inheritance

Prestige does not guarantee current review

Museum or archive records may preserve donor statements, legacy terminology, old measurements or attributions that have never been rechecked.

Misrepresentation

Some evidence may be altered or invented

Forged receipts, false labels, manipulated photographs, invented provenance and altered serial numbers require stronger verification, especially where value or title is affected.

Method

A rigorous conflict-resolution workflow

The sequence below moves from definition to evidence testing, conclusion and preservation. It is deliberately slower than choosing the most prestigious or recent-looking source because those shortcuts often preserve the original error.

State the disputed proposition precisely

Replace 'the sources disagree' with a sentence that identifies the object, attribute, territory, event, date and type of claim.

For example: Source A states that this boxed game was first commercially released in the UK in 1981, while Source B assigns the first UK retail release to 1982.

Separate the claims

Break broad statements into smaller propositions: announced, copyrighted, manufactured, shipped, sold, acquired or later catalogued. The conflict may disappear when different events are no longer collapsed into one date.

What exactly is each source dating or identifying?
Does a statement concern the product family, a variant or this individual object?
Are manufacture, publication, distribution and purchase being confused?

Verify each source and preserve a usable reference

Record the full title, author or issuing body, date, edition or revision, page or record number, archive reference, stable link or saved copy, and access date. A claim that cannot be relocated is difficult to test or revise later.

Map the source genealogy

Identify whether each source is original, derivative or of unknown dependence. Trace citations, copied wording, shared photographs and imported database records back to the earliest accessible assertion.

Do not resolve a conflict by counting repetitions. Five derivative entries may provide only one underlying claim.

Ask what each source was in a position to know

A factory ledger may establish an order date, quantity or supplier but not retail availability. An owner's recollection may establish acquisition and accompanying contents at that time, but not original factory configuration.

Assess purpose, incentive and proximity

Consider why the source was created and its temporal, physical, organisational and informational proximity to the disputed fact. Proximity helps only where the source creator could actually observe the relevant matter.

Check internal consistency

Look for mismatched dates, impossible ownership overlaps, dimensions inconsistent with photographs, product codes belonging to another object, or serial formats that post-date the claimed period.

Examine external and object-based evidence

Compare the claims with verified examples, archival documents, period advertisements, invoices, correspondence, dated photographs and institutional records. Then inspect the object independently, recording observations before reading them through either disputed claim.

Generate alternative explanations

Do not force the case into Source A versus Source B. Both may be partly right; they may describe different variants, stages, objects or components; one may copy an error; the object may be composite; or the available evidence may not resolve the question.

Seek discriminating evidence

Prioritise evidence that could separate the live hypotheses. If the question is announcement in 1981 versus retail release in 1982, dated invoices, distributor shipping records, reviews or customer receipts are more useful than another undated guide.

Reach a conclusion and assign confidence

Use outcomes such as resolved, provisionally resolved, partly resolved, unresolved or unresolvable with current evidence. State confidence separately as high, moderate, low or indeterminate.

Preserve the disagreement and revision history

Keep the competing claims, why one interpretation was preferred, what evidence was decisive, what remains uncertain, what could change the conclusion, and who made the assessment and when.

Judgement

How to weigh competing evidence

No numerical score can mechanically resolve every case. These dimensions structure judgement and expose why one claim is stronger for the particular fact under review.

Directness

Observed, inferred or copied?

Direct observation generally carries more weight for the feature observed, but interpretation and object integrity still matter.

Relevance

Does it answer the exact question?

A source may be authoritative yet only indirectly related to the disputed attribute.

Independence

Is it genuinely separate?

Repeated claims derived from one source should not be counted as multiple confirmations.

Contemporaneity

How close is it to the event?

A contemporary source may be useful but can still describe plans, promotion or incomplete knowledge.

Competence

Did the creator have relevant knowledge?

Expertise must match the claim: manufacturing, bibliography, provenance and material analysis require different competencies.

Specificity

This object or a general type?

Evidence tied to the exact object, variant or transaction is usually stronger than broad category statements.

Verifiability

Can the evidence and reasoning be checked?

Accessible records, photographs, measurements and methods allow later researchers to reproduce the assessment.

Consistency

Does it fit internally and externally?

Contradictions within a source or with stronger evidence reduce the force of the relevant claim.

Incentive

Was there reason to exaggerate or simplify?

Commercial, reputational, legal or institutional pressures affect weighting without automatically invalidating the source.

Chain integrity

How many undocumented steps intervene?

Each unrecorded transfer, paraphrase or image reuse increases the risk that meaning has changed.

Evidence / meaning / collector risk

Evidence

A mould mark is directly visible and consistently associated with Company B on verified examples.

Meaning

It may indicate manufacture by Company B, but not necessarily design, publishing, branding or distribution.

Collector risk

Recording 'made by Company B' without qualification may create a false attribution and affect variant identity or value.

Confidence

Use calibrated conclusion language

High confidence

The surviving evidence establishes...

Use where independent, relevant and verifiable evidence consistently supports the same conclusion and serious alternatives have been tested.

Moderate confidence

The evidence indicates...

Use where one interpretation fits the evidence best but a material gap or plausible alternative remains.

Low confidence

This may represent...

Use where evidence is limited, derivative, ambiguous or dependent on assumptions that have not been tested.

Unresolved

Available sources conflict

Use where current evidence cannot distinguish between live explanations. Record the evidence needed rather than manufacturing precision.

Application

Collector scenarios

Scenarios make the method concrete, but the same principle applies across categories: define the claim, separate layers, inspect the evidence and avoid turning a plausible explanation into a proven fact.

Conflicting first-edition claims

Situation: A dealer calls a book a first edition; a bibliography calls it a second issue.

Investigate

Separate edition, printing, issue, state, binding, territory and publication sequence.
Inspect title and copyright pages, printer's key, binding points and advertisements.
Check the bibliography's definitions and evidence rather than comparing labels alone.

Collector judgement: The sources may not be using 'first edition' in the same bibliographical sense. Record the physical points first and apply the relevant terminology explicitly.

Different component lists

Situation: A guide lists twelve pieces; three apparently unopened examples contain eleven.

Investigate

Check whether the guide used a pre-release or regional list.
Test whether one component was optional, paired or introduced later.
Verify that the examples are genuinely unopened and belong to the same production period.

Collector judgement: Do not declare the guide or objects wrong until territory, variant and production sequence are defined. The conflict may reveal a contents change.

Conflicting auction prices

Situation: One database records $2,000 and another $2,400 for the same lot.

Investigate

Identify hammer price, buyer's premium, tax and currency treatment.
Verify that the lot sold and that both records describe the same composition.
Distinguish estimate, asking price and realised price.

Collector judgement: Both figures may be accurate representations of different price concepts. Normalise before using either as valuation evidence.

Provenance disagreement

Situation: An auction catalogue says 'private collection, London, acquired 1975'; a family letter suggests ownership in 1968.

Investigate

Determine whether 1975 refers to the current owner rather than the whole family chain.
Confirm that the letter identifies the exact object.
Check for transfers within the family or two similar objects.

Collector judgement: Compressed provenance wording should not be read as a complete chain. Preserve both records and state the unresolved interval precisely.

Manufacturer attribution conflict

Situation: Packaging names Company A while a mould mark is associated with Company B.

Investigate

Separate design, manufacture, publishing, branding and distribution.
Consider subcontracting, mould transfer, replacement components and batch changes.
Compare known examples from the same territory and period.

Collector judgement: 'Made by' is too broad. The evidence may support Company A as brand or publisher and Company B as component manufacturer.

Warning signs

When a conflict deserves greater caution

A high-value conclusion depends on one undocumented assertion.
The seller supplies both the provenance and the only authentication.
Many sources repeat identical wording, spelling errors, images or unexplained dates.
Claims become more specific as they are copied, without new evidence.
The decisive feature is not shown in photographs or described in measurements.
Contrary evidence is dismissed by prestige, ridicule or assertion rather than analysis.
Exact production or survival totals appear without records or a stated counting method.
A database entry changes without a visible revision history.
A famous ownership association rests on an incomplete or ambiguous chain.
The decisive evidence is said to be confidential and cannot be described sufficiently for review.

Action

Match the response to the stakes

A minor descriptive disagreement does not need the same response as a disputed signature, ownership claim or suspected forgery. Escalate proportionately while preserving evidence at every level.

Level 1

Low-stakes descriptive conflict

Example: Minor colour terminology or subjective finish description.

Preserve both descriptions
Photograph under controlled conditions
Use defined terminology
Note subjective variation

Level 2

Classification conflict

Example: First versus second printing, regional issue or component state.

Define the disputed terms
Compare physical points
Trace source dependency
Assign a provisional status

Level 3

Financially consequential conflict

Example: Rare variant, signature attribution or value-changing completeness claim.

Seek independent examination
Do not rely on seller evidence alone
Retain all documentation
Disclose uncertainty in sale or valuation records

Level 4

Authenticity, title or ethical conflict

Example: Suspected forgery, stolen property, unlawful export or high-risk provenance gap.

Pause acquisition or disposal
Preserve evidence and communications
Seek qualified specialist and legal advice where appropriate
Avoid unsupported public accusations

Domain boundary

Conflict analysis does not replace specialist authentication, provenance or legal work

Source evaluation can clarify the disagreement and improve the evidence record. It cannot by itself authenticate a signature, establish lawful title, perform material analysis or resolve complex export and cultural-property questions.

Where the disputed issue crosses those boundaries, use the conflict record to define the question and brief the appropriate specialist rather than strengthening the wording beyond the available evidence.

Documentation

Build a durable conflict record

A durable record allows another collector to reconstruct the disagreement, test the reasoning and understand later revisions. Attach the conflict to the disputed claim or attribute rather than labelling the whole object as unreliable.

Record field	What to capture
Disputed attribute	Date, maker, variant, provenance, completeness, sale status or another defined attribute.
Exact question	The precise proposition being tested, including object, territory, event and date where relevant.
Competing claims	Accurate paraphrases or short quotations, each kept separate.
Source references	Full title, creator, date, edition, page, lot, archive reference, link or saved copy.
Object observations	Measurements, marks, materials, construction, component relationships, wear and alterations.
Source relationship	Independent, derivative, copied, shared evidence or unknown.
Definitions	How each source uses disputed terms such as edition, issue, complete or original.
Evidence assessment	Strengths, limitations, purpose, proximity, competence, consistency and incentives.
Preferred interpretation	Current conclusion or explicitly unresolved alternatives.
Confidence	High, moderate, low or indeterminate, stated separately from the conclusion.
Remaining gap	The evidence that would most usefully discriminate between the alternatives.
Revision history	Researcher, date, later changes and the reason each conclusion was revised.

Suggested status vocabulary

No known conflictApparent conflict - definitions differApparent conflict - different objectsActive conflictPartly reconciledProvisionally resolvedResolvedUnresolvable with current evidenceSuperseded claim retainedRequires specialist review

Specialist threshold

When to stop researching alone

High financial value or major catalogue significance
Suspected forgery, altered evidence or disputed signature
Scientific dating, paper, ink, pigment, fibre or manufacturing analysis
Export, ownership, cultural-property, wartime or colonial provenance questions
Complex archival handwriting or specialist production records
Conservation evidence that may be damaged by further examination
Potentially defamatory allegations or contested legal title

A specialist opinion should still identify the evidence examined, scope of expertise, method, whether the object was seen in person, limitations, level of certainty and date of assessment. Scientific results may establish composition, process or an approximate date range without proving exact maker, ownership, edition or original assembly.

Decision test

Ten questions before you close the record

Are the sources answering the same question?
Do they refer to the same object, issue, variant, component or event?
Are their key terms defined in the same way?
What evidence lies behind each claim?
Are the sources independent?
What was each source in a position to know?
What does the physical object show before interpretation?
Which explanation fits the strongest evidence with the fewest unsupported assumptions?
What evidence would distinguish the remaining alternatives?
How certain can the conclusion honestly be?

Productive outcome

A conflict can be the beginning of discovery

Careful conflict analysis can reveal unrecognised variants, transitional states, regional editions, reused packaging, catalogue-number changes, component substitutions, false rarity traditions, misidentified auction records and corrected provenance chains. The goal is not merely to choose a winner, but to improve the object history.

Key takeaways

A conflict is between particular claims, not automatically between whole sources.
Many apparent contradictions disappear when dates, terms, objects, variants, territories and components are defined precisely.
Sources should be weighted by relevance, independence, access, competence, purpose and verifiability - not counted or ranked by prestige alone.
Object evidence is powerful but must be interpreted in light of replacement, restoration, composite assembly and production variation.
A responsible record preserves competing claims, the reasoning behind the current conclusion, its confidence and the evidence that could change it.

Continue learning

Corroboration and Independence

Test whether apparent agreement comes from independent evidence or repeated dependence on one underlying source.

→

Back to Source Evaluation

Return to the Source Evaluation sub-domain and its full sequence of topics.

→

Source Currency and Revision

Continue to judging when a source was created, updated, superseded or silently inherited from earlier work.

→