People by country: six different questions hiding behind one label

"Country" looks like the simplest field in a biographical record. It is one of the least stable. Most reference works carry a single country column, most of them have quietly picked one of six incompatible meanings for it, and almost none of them tell you which.

What to take away

Six questions wearing one label

Ask a person's country and you could be asking any of these:

These answers diverge constantly. Someone born in a port city, educated in a second country, naturalized in a third, and buried in a fourth has four true country answers and no single correct one. A database that stores one value has made an editorial choice. Your job is to find out which choice, before you repeat it.

Borders move; the people under them do not have to

Empires dissolved. Territory changed hands after wars. Colonies became independent states. Federations split, and a few merged. For an enormous number of people in the historical record, the place of birth is still on the map and the country of birth is not.

The change history published with ISO 3166-1 is a short course in how often it happened. The same problem at state level is set out on the US state pillar, and at municipal level on the city pillar.

That produces two defensible conventions:

Both are used by serious reference works. Neither is wrong. What is wrong is switching between them inside one dataset, which is exactly what happens when records are merged from several sources.

The practical fix is to stop treating this as one field. Record the settlement, then the governing polity, then the date the polity applied. Place names outlive states, and a place with a date is recoverable no matter how many times the flag changes.

Watch for the harder cases. Contested territory can leave two states each maintaining an official record. A city whose name changed with its government appears under two spellings that no string match will connect. A region administered by one state and legally annexed by another years apart gives you two different "correct" dates for the same change.

Nationality, citizenship, and birthplace are not synonyms

They get used interchangeably in prose and they answer different questions.

Birthplace is a geographic fact about a single event. It says nothing about where a person was raised, what language they spoke, or which passport they held. A birth in transit, in a hospital in the nearest large town, on a military installation, or during a posting abroad detaches birthplace from everything readers assume it implies.

Citizenship is a legal status. It is conferred, inherited, applied for, renounced, revoked, and sometimes held in plural. It has a start date and can have an end date.

Nationality is the ambiguous one. In legal drafting it usually means citizenship. In ordinary speech, and in a great deal of older scholarship, it means membership of a national or ethnic community that may have no state at all. A person can belong to a nation with no country and hold the citizenship of a country whose national identity they do not share. Both statements can be recorded, but not in the same column.

Residence is where someone actually lived, which is often the most informative field and the one least likely to be recorded.

Citizenship at birth cannot be inferred from a birthplace

States assign citizenship at birth on two broad principles: by place of birth, and by descent from a parent. Most states use some blend, the blend has been amended repeatedly, and amendments have sometimes applied retroactively to people already born.

So the citizenship a person held at birth is a function of the law in force in that place at that time, plus their parents' status. In some periods it also turns on their parents' marital status and which parent held what. You cannot derive it from a map.

If a claim about citizenship matters to what you are writing, find the nationality statute of that state for that period, or a legal history that cites it, and record what you found. If you cannot find it, write that the citizenship is not established rather than assuming the obvious answer.

The same caution applies at the other end. Naturalization, loss of citizenship on marriage under older laws, statelessness after a state ceased to exist, and posthumous or symbolic grants of citizenship all put a date on a status that databases usually store without one.

What the catalogs actually store

It is worth being precise about the sources most often cited for this, because they are narrower than they look.

National library name authority records exist to disambiguate names for cataloging, and the same authority control practice runs through museum and archive vocabularies. A record establishes one authorized form of a name, lists variant forms, and often adds dates and a fuller form of the name. It is a control device for finding books, not a biography. Its purpose is met if two people who share a name can be told apart.

The Virtual International Authority File clusters authority records from many national libraries so that the same person's records in different countries link together. The clustering is largely automated. Two clusters can describe one person; one cluster can absorb two people with similar names and dates. A VIAF identifier is a strong hint, not a proof of identity.

Wikidata is closer to what people want, because it keeps place of birth and country of citizenship as separate properties and allows qualifiers that put start and end dates on a citizenship. That structure is good. The weakness is that any individual statement may carry a solid reference, a reference to another tertiary summary, or no reference at all. Check the statement, not the item.

The general point: the presence of a person in a national catalog is a fact about that country's libraries. It tells you something about collecting and publishing. It does not tell you where the person was from.

Institutions assign countries by their own rules

A large share of the country labels attached to people did not come from a passport or a birth certificate. They came from a body that had to decide who counts as representing where, and wrote a rule.

Sports federations set eligibility by some combination of citizenship, birth, parental or grandparental descent, and a minimum period of residence, and they change those rules. Prize and award bodies restrict entry by nationality, by residence, by place of publication, or by language, and each of those produces a different answer.

National academies, delegations, and funding schemes have their own membership conditions. Publishing and broadcasting quotas define national origin by where a work was produced rather than by where anyone was born.

Two things follow. First, different institutions can correctly record the same person under different countries, because each uses a different test and none claims to describe identity.

Second, a dataset built from award records, team rosters, or membership lists inherits whichever tests those bodies used, and those tests are usually documented somewhere the dataset does not mention.

When you take a country from a source of this kind, record the rule with it: represented a country under a federation's eligibility criteria, or was eligible for a prize restricted to a particular nationality. That is what the record supports.

It is not the same sentence as a statement about where someone is from. Readers will treat it as one unless you write the difference in.

The bias you will mistake for a finding

Count records per country in any large biographical dataset and you will get a ranking. The ranking will look meaningful. It mostly measures:

None of that is a property of the people. A long-running, well-funded national library will outrank a country with an older and larger literary tradition whose records were destroyed or never centrally indexed.

Publishing that ranking as a finding about nations is the single most common error in this subject. It is why this site does not publish country league tables of people.

Recording a country claim so it stays true

Living people

Do not state a living person's nationality, citizenship, ancestry, or birthplace on the strength of a database entry, a surname, a place of residence, or a photograph. People describe their own origins in ways no schema captures, legal status changes without announcement, and a confident wrong answer about a living person is both false and insulting.

If the question genuinely has to be addressed, the only usable evidence is what the person has said on the record, or an official document they have made public. Attribute it: report that they described themselves a particular way, and when. That sentence stays true even if the underlying situation changes. A flat assertion does not.

Questions to ask of any country field

Bottom line

A country field compresses a claim about geography, law, and time into one word, so decompress it before use: what basis, what date, what polity, what source.

Related material on this site is organized by heritage, by period, and by work, and the same discipline applies to all three. The country looks easiest and is not.

Common questions

Which convention should a country field use, contemporary or modern?

Either, provided the dataset uses one of them everywhere and records which. The damage is done by switching, which happens invisibly whenever sources are merged.

Is a name authority record evidence of nationality?

No. It is a cataloging device for telling two similarly named people apart, and any place in it usually comes from a title page rather than from research.

How should a disputed or unrecognized entity be stored?

As the entity the record names, with a note on its status and the date. Silently mapping it to a recognized neighbor changes what the record said.

What belongs in the field when a person held more than one citizenship?

All of them, each with its date and its document. A single value forces a choice the sources did not make.