The Document That Contradicts the Record

A conflict between two official documents feels like a discovery of wrongdoing and almost never is. Registers are filled in by people working from what they were told, and the same event recorded twice by two offices produces two accounts. What matters is not which document is authentic, since both usually are, but which is closer to what happened.

Why records disagree in the first place

Different informants. A birth is registered by whoever attends the office, and a passport is applied for by somebody else years later. Two people reported the same event and neither was checking the other's work.

Different purposes. A hospital note is written to treat a patient and a certificate is written to create a legal record. Each is careful about the things its own purpose requires and careless about the rest.

Transcription between systems. Every time a particular is copied from one register to another there is an opportunity for a digit to move or a name to acquire a letter, and large systems copy constantly.

Changes that were never propagated. A correction made in one register rarely reaches the others, so an amended entry and an unamended copy of it can coexist quite happily for decades.

Deliberate misstatement, occasionally. It happens, and it is the least common explanation. Treating it as the first explanation wastes time and antagonizes decision-makers who see the ordinary version constantly.

What decides which record prevails

Proximity to the event. A record made on the day, by somebody present, generally beats one made years later by somebody working from a form. This single factor resolves most conflicts without any further argument.

The maker's actual knowledge. A registrar records what an informant states. Where the informant was a grandparent recalling a date, the entry has the reliability of that recollection and no more, whatever the seal on the certificate says.

Whether accuracy mattered at the time. Entries made where a mistake would have been noticed and corrected carry more weight than entries on a form nobody read again, which is the reasoning behind the treatment of routine business entries.

Independence. Two records that both descend from a single original statement are one piece of evidence appearing twice, and counting them as two is the commonest error in this area.

Internal consistency. A document that agrees with itself on ages, intervals and relationships is more likely to be right than one that does not, and internal arithmetic is worth checking before anything else.

The corroboration trap

Copied records look like agreement. Five documents carrying the same wrong spelling of a name look like overwhelming corroboration until somebody traces all five back to the same application form.

Trace each document to its source. The question is always where the particular came from, not how many places it now appears in, and that question is answerable from the documents themselves.

Later records inherit earlier errors. Institutions copy forward, and an error introduced early travels quietly into every record made afterward, gathering apparent authority as it goes.

Independent sources are what count. A school register and a baptismal entry recording the same date, from different informants, are worth more than six later certificates that all descend from one another.

Say so explicitly. Where the other side's documents are all copies of one another, pointing that out plainly is usually more effective than attacking any one of them.

Resolving the conflict

Find the earliest record. Work backward through each chain until the first appearance of the disputed particular, because that is the only entry whose accuracy anybody can actually assess.

Identify who supplied the information. Registers frequently name the informant and their relationship to the subject, which converts an anonymous entry into a statement by an identifiable person.

Look for a contemporaneous check. Anything made at the time for a different purpose — a payroll record, a school admission, a ship's manifest — is powerful precisely because nobody was thinking about this dispute.

Apply to correct the record. Most registers have a correction procedure, and a corrected entry removes the conflict permanently rather than requiring it to be argued in every future proceeding.

Consider that both may be wrong. Two documents disagreeing does not mean one of them is right, and reconstruction from fragments is sometimes the only honest route.

Presenting an unresolved conflict

Produce both documents. Concealing the unhelpful one converts an ordinary evidential problem into a credibility problem, and the other side will produce it anyway.

Say which is relied on. A statement that identifies the preferred record and gives the reason is far stronger than one that quietly uses a particular without explaining where it came from.

Give the reason in terms of reliability. Proximity, knowledge, purpose and independence are the grounds a decision-maker recognizes, and an argument framed in those terms lands better than an assertion.

Anticipate the obvious question. If the preferred document is the later one, the reason it is preferred needs stating explicitly, because the presumption runs the other way.

Accept the residual uncertainty. A conclusion offered as probable is more credible than one offered as certain, particularly where the documents plainly disagree, as explaining a gap honestly sets out.

The habit worth building is genealogical rather than adversarial. Every particular in every document came from somewhere, and tracing it back is a mechanical exercise that resolves conflicts far more often than arguing about which certificate looks more official.

It also protects against a trap that catches careful people: assembling a large volume of consistent documentation and presenting it as overwhelming, only to have somebody point out that all of it descends from a single form completed by a relative who was guessing.

Where the conflict cannot be resolved, the correction procedures maintained by most registers are underused. They are slow and they require evidence, but a corrected entry ends the problem for every future purpose rather than for this one proceeding.

Above all, the discovery that two official documents disagree is not a discovery about anybody's honesty. Records are made by people working quickly from what they are told, and the wonder is that they agree as often as they do.