Resolving Conflicting Evidence in Family History Research

Conflicting evidence is an inherent feature of genealogical research, not an anomaly. When two or more records provide incompatible information about the same event—such as differing birth dates on a census return and a baptismal register—the researcher faces a structured analytical task governed by the Genealogical Proof Standard. This page describes the professional framework, classification systems, and resolution mechanics that the genealogical community uses to handle evidentiary conflicts, and maps the boundaries between resolvable and irresolvable contradictions.

• Definition and scope
• Core mechanics or structure
• Causal relationships or drivers
• Classification boundaries
• Tradeoffs and tensions
• Common misconceptions
• Checklist or steps (non-advisory)
• Reference table or matrix
• References

Definition and scope

Conflicting evidence in genealogical research refers to the presence of two or more pieces of information, derived from separate or even the same sources, that cannot simultaneously be true about a single genealogical assertion. The Board for Certification of Genealogists (BCG) defines a genealogical conclusion as proven only when all five elements of the Genealogical Proof Standard (GPS) are satisfied, one of which explicitly requires "resolution of conflicting evidence" (BCG, Genealogical Standards Manual). Without resolution, no assertion—regardless of how much supporting evidence exists—meets the professional standard of proof.

The scope of evidentiary conflict extends across every record type used in the discipline: vital records, census returns, church and parish records, probate files, military service files, immigration documents, and DNA test results. Conflicts are not limited to factual assertions like dates and places; they also arise in identity questions (whether two records refer to the same individual), relationship claims (parentage, siblingship), and event interpretations (whether a land transfer was a sale or an inheritance). The conceptual overview of family history research frames evidence analysis as a core operational function rather than a peripheral task.

Core mechanics or structure

Resolution of conflicting evidence operates through a defined sequence of analytical steps codified by the BCG and reflected in the credential requirements of professional genealogical organizations such as the Association of Professional Genealogists (APG) and the International Commission for the Accreditation of Professional Genealogists (ICAPGen).

Source classification. Every source is classified as original (the first recording of information) or derivative (a copy, abstract, transcription, or index of the original). A derivative source carries higher risk of introduced error. The 1850–1940 U.S. federal census enumerator sheets are original sources; Ancestry.com index entries for those sheets are derivative.

Information classification. Information within a source is classified as primary (provided by a knowledgeable participant or eyewitness), secondary (provided by someone reporting secondhand), or indeterminate (where the informant cannot be identified). A death certificate typically contains primary information about the date and place of death (recorded by the attending physician) but secondary information about the decedent's birth date and birthplace (reported by a surviving family member).

Evidence classification. Evidence is either direct (the source explicitly states the fact in question) or indirect (the fact must be inferred from combining the source with other information). A marriage license directly evidences a marriage; a census listing of a child with both parents at an address indirectly evidences the marriage.

When two pieces of direct evidence conflict—for instance, a birth certificate stating 14 March 1873 and a baptismal record stating 12 March 1873—the platform evaluates the informant proximity, time elapsed between event and recording, and the institutional incentives behind each record. Resolution is documented in a written proof argument or proof summary, as specified by GPS Element 5. The written analysis must explicitly acknowledge the conflicting item and explain, with cited reasoning, why one piece of evidence is given greater weight. Source citation at the individual-assertion level is mandatory throughout this process.

Causal relationships or drivers

Conflicting evidence does not appear randomly. Specific, identifiable causes produce predictable categories of conflict.

Informant error and secondary reporting. The most frequent cause of conflicting dates and places is reliance on informants who did not witness the event. A 2014 analysis by researcher Thomas W. Jones, published in Mastering Genealogical Proof (Arlington, VA: National Genealogical Society), demonstrated that death certificates routinely disagree with birth registrations on birth dates because the death certificate informant (often a spouse or adult child) recalled incorrectly.

Clerical and transcription error. Handwritten documents introduce transcription drift. The 1880 U.S. Census employed approximately 31,382 enumerators (U.S. Census Bureau, History: 1880 Census), each with individual penmanship. Misreading a numeral (3 vs. 5, 7 vs. 1) and misspelling surnames—addressed more fully at the Soundex and name variation page—produce records that appear to describe different people.

Deliberate misrepresentation. Age falsification for military enlistment (especially during the Civil War and World War I), immigration fraud involving altered names or nationalities, and concealed parentage for social or legal reasons generate conflicts that are not errors but intentional distortions. Unknown parentage research and adoption research regularly encounter deliberate record manipulation.

Systemic record variation. Prior to the adoption of the Gregorian calendar in British colonies in September 1752, dual dating (Old Style / New Style) created legitimate two-date entries for a single event. The historical dates and calendars reference covers this systemic driver. Likewise, geographic name changes produce location conflicts when jurisdictions shifted (e.g., Virginia counties that became West Virginia in 1863).

DNA-record discordance. DNA evidence may conflict with documentary evidence by revealing non-paternity events, adoptions, or misidentified parentage. According to a frequently cited 1999 meta-analysis by Mark A. Bellis and colleagues published in the Journal of Epidemiology and Community Health (Vol. 59, pp. 749–754, 2005 revision), the median non-paternity rate across studied populations is approximately 3.7%, though rates vary substantially by population. The autosomal DNA versus Y-DNA versus mitochondrial DNA page addresses which DNA test types reveal which kinds of discordance.

Classification boundaries

Not every disagreement between records constitutes a genuine conflict requiring GPS-level resolution. The discipline distinguishes among three categories:

Apparent conflict. Two records appear to disagree but are reconcilable without formal proof argument. Example: a birth date of "March 1873" in one source and "14 March 1873" in another. The first is less specific but not incompatible. No resolution argument is needed—only a note acknowledging the specificity difference.

True conflict. Two records provide mutually exclusive assertions. Example: a birth date of 14 March 1873 versus 14 September 1873. Both cannot be correct. A written proof argument is required.

Irresolvable conflict. After exhaustive search, the evidence remains evenly balanced or the conflicting sources are of equal reliability and proximity. The GPS requires that the researcher acknowledge the irresolvable state rather than force a conclusion. This does not mean the research failed—it means the extant record base is insufficient to resolve the question. The distinction between "not yet resolved" and "irresolvable given surviving records" is operational, not theoretical.

A fourth boundary distinguishes evidence conflicts from interpretation conflicts. Two researchers may disagree on the meaning of indirect evidence (e.g., whether a man listed in a household is a boarder or a son-in-law). This is an inferential dispute, not an evidentiary conflict in the strict GPS sense, though it follows the same resolution mechanics.

Tradeoffs and tensions

Depth versus closure. Exhaustive search—GPS Element 1—has no bright-line endpoint. The tension between thoroughgoing research and the practical need to reach conclusions is real. Professional genealogists operating under client engagement terms face time and cost constraints that academic researchers do not. The BCG's standard requires a "reasonably exhaustive search" rather than absolute exhaustion, but the boundary is subjective and debated within the certification community.

Quantitative DNA evidence versus qualitative documentary evidence. DNA results provide probabilistic matches measured in centimorgans (cM), while documentary evidence provides narrative context. When a 45 cM shared segment suggests a 4th-to-6th cousin relationship but documentary evidence suggests no connection, the researcher must reconcile fundamentally different evidentiary paradigms. Weighting one type over another involves discipline-specific judgment that professional organizations have not fully standardized.

Original versus accessible. The preference for original sources over derivative ones creates a practical tension: original records may be held in repositories with restricted access (National Archives and Records Administration, state archives), while derivative versions (indexed databases, digitized abstracts) are freely or cheaply available. The tradeoff between using the best available derivative and traveling to consult the original is a resource allocation decision built into every conflict resolution.

Narrative coherence versus evidentiary atomism. Writing a family history narrative requires presenting conclusions. Overemphasis on unresolved conflicts can paralyze narrative construction; underemphasis can produce a narrative that presents contested facts as settled. The discipline's answer—explicit proof arguments embedded in or appended to narrative text—adds substantial labor.

Common misconceptions

"The oldest record is always the most reliable." Age of a record correlates with proximity to the event only if the record was also created near the time of the event. A family Bible entry might have been inscribed decades after the events it records, making it secondary information in an original source—potentially less reliable than a later-created but contemporaneous record such as a church register entry made at the time of baptism.

"Two records agreeing proves the fact." Agreement between records does not constitute proof if both records trace to the same informant. A death certificate and a published obituary may both report the same incorrect birth date because the obituary writer obtained the date from the same family member who informed the death certificate. This is dependent-source agreement, not independent corroboration.

"DNA evidence overrides all documentary evidence." DNA results are powerful but probabilistic. A shared cM value of 35 cM, for instance, is consistent with relationships ranging from 3rd cousin to half-5th cousin. DNA cannot, by itself, specify the exact genealogical pathway. Documentary evidence and DNA are complementary, not hierarchical.

"Conflicts mean one record is wrong." Both records may contain partial truths. A person recorded as born in "Virginia" in 1870 and "West Virginia" in 1880 may be reflecting a jurisdictional change, not a factual error.

"Indexing databases are sufficient for exhaustive search." Indexes are derivative sources subject to transcription error, OCR misreading, and selective inclusion. Relying solely on indexed databases (even well-constructed ones) does not satisfy the exhaustive-search requirement of the GPS. Consulting original images, visiting repositories, and reviewing land records, city directories and voter rolls, and newspapers in unindexed collections may be necessary.

Checklist or steps (non-advisory)

The following sequence reflects the GPS-compliant resolution process as outlined by the BCG and the National Genealogical Society:

1. Identify the specific assertion in conflict. Isolate the exact claim (date, place, name, relationship) where records disagree.
2. Classify each source. Determine whether each source is original or derivative using the framework from understanding genealogical records.
3. Classify the information. Determine whether each piece of information is primary, secondary, or indeterminate by identifying the informant.
4. Classify the evidence. Determine whether each piece of evidence is direct or indirect relative to the assertion.
5. Evaluate informant proximity. Assess each informant's relationship to the event, the elapsed time between event and recording, and the informant's possible motivations.
6. Search for additional sources. Conduct further research in oral history, photographs and heirlooms, school records, cemetery records, Social Security Death Index, and passenger lists to locate corroborating or clarifying evidence.
7. Construct a correlation table. Map each source against the assertion, recording what each says, its classification, and its assessed reliability.
8. Draft a written proof argument. Explain the conflict, the evidence on each side, and the reasoning for the conclusion. Cite every source at the assertion level using proper citation standards.
9. Record the conclusion or acknowledge irresolvability. Assign a status: resolved (with stated conclusion), provisionally resolved (pending further research), or irresolvable (with explanation). Incorporate findings into family group sheets and family tree structures.

Reference table or matrix

The following matrix maps common conflict types to their typical causes, resolution strategies, and relevant record types. The genealogy authority home page provides further orientation to these record categories.