People by country directory: facts, examples and context

Two kinds of reference work stand behind almost every statement that a person "was" of a particular country. One is the national biographical dictionary. The other is the name authority file. They were built for opposite reasons, and confusing them is the most common way a cataloging convenience turns into a biographical claim.

What to take away

• A national biographical dictionary applies an inclusion rule chosen by its editors. Its country boundary is editorial, and the rule is usually printed in the preface nobody reads.
• A name authority record exists to tell two people apart. Any country in it is a disambiguation aid drawn from cataloged works, not an assertion about nationality.
• Aggregated authority data inherits every wrong merge in its inputs, and a wrong merge transfers attributes between two real people.

What a national biographical dictionary is

It is a bounded editorial project. Someone decided which lives belong to the nation and which of those are significant enough to include, then commissioned entries against those decisions.

Both decisions vary between projects, and they vary in ways that change who appears:

• Some works admit people born in the territory. Some admit people who worked there. Some admit people who shaped the national culture from outside it, and exclude people born there who did not.
• Significance thresholds differ by field. A project may cover public office exhaustively while covering trades barely at all.
• Coverage of women, of working people, and of communities the compilers were not connected to is thinner in older works, and later editions add entries rather than replacing the shape of the original.
• Revision is continuous in some projects and frozen in others. An absence in an old edition is evidence about the edition.

None of that makes these works unreliable. They are frequently the best available account of a life, written by someone who read the sources. It makes them works with a scope, and the scope has to travel with anything you take from them.

So cite the work, the edition, the entry and the date, and read its statement of scope before you quote its country. Where an entry says a person belongs to the nation, that is the project's rule speaking, and the rule can be checked.

What a name authority record is for

Libraries need to know whether the person on this title page is the person on that one. A name authority record answers that: it fixes a preferred form of a name, lists variant forms, and attaches whatever attributes help distinguish this person from others with similar names, such as approximate dates or an associated place or field.

Those attributes are chosen for their power to disambiguate, and they come from the works being cataloged rather than from research into the person. A place appearing in an authority record may mean the person published there, was described there, or appeared on a title page there. It is not the outcome of an investigation into nationality, and it was never meant to be read as one. The disambiguation purpose is visible in the fields themselves in any published vocabulary, such as the Getty's Union List of Artist Names.

The distinction is easy to state and constantly ignored: cataloged as associated with a place is not the same claim as was a national of a state.

What aggregation does to both

Authority files from many institutions are matched and clustered so that one person has one cluster. Clustering is done at scale by algorithm, with human correction after the fact, which is the practical limit of authority control once it leaves a single catalog.

Two failure modes follow, and both propagate:

Wrong merges. Two people with similar names and overlapping dates become one record carrying both sets of attributes. The merged record now asserts two countries, two fields of work and an implausible span, and downstream consumers pick one of each.

Wrong splits. One person cataloged under a maiden name, a married name and a transliteration becomes three records, each with partial attributes, and each looks like a thin biography.

Machine-readable knowledge bases pull from these clusters, and content pipelines pull from the knowledge bases. By the time a country appears on a profile page, several layers have converted a cataloging aid into an attribute, and no layer has added evidence. The general habit of walking a claim back to the object that supports it is set out on the historical records page.

What a responsible directory entry contains

If you are building one, the useful shape puts the evidence in the row rather than under it:

Element	What goes in it
Statement	The specific relation, for example naturalized in a named state on a date
Basis	Birth, citizenship, residence, work, or self-description
Source	The work, edition and page, or the record and its repository
Scope note	The inclusion rule of the source, where it has one
Date	When the relation held, and when the source was consulted
Status	Established, disputed between sources, or unresolved

A directory built that way holds fewer entries and each one survives being checked. A directory built from a single flat country field holds more and none of them do.

What we do not publish

A list of people by country. Assigning one country per person means picking one relation among several and discarding the rest, which the country pillar sets out at length. Published as a directory, that editorial pick reads as a fact about the person.

Any ranked tally by origin. No counts of which country produced the most of anything. Such a count is dominated by which national institutions funded biographical projects, which archives were digitized, and which language the aggregator indexes best. It orders reference works, not people, and even a clean count would be an exercise in scoring populations against each other that we decline to run.

Claims about living people. No living person's nationality, ancestry, religion or birthplace is asserted here, including where an authority record or a knowledge base supplies one. Where a person has publicly described their own background, we quote and date it. The reasoning is in the heritage guide, and the code and naming problems underneath the field itself are on the country list page.

Bottom line

Biographical dictionaries tell you who their editors thought belonged to a nation. Authority files tell you how librarians told two names apart. Neither was built to answer the question people now ask of them, and the answer they appear to give has been through three layers of copying by the time it reaches a profile.

Common questions

A knowledge base and a biographical dictionary disagree. Which do I follow?

Neither, until you know where each got its value. The dictionary usually has an author and a source note; the knowledge base usually has an import. Follow both back and cite the one that actually has evidence, recording the disagreement rather than hiding it.

Is an authority record ever the right citation for a country?

For the fact that a cataloging body associated a name with a place, yes, and say it that way. For a person's nationality, no. There is no cataloging standard under which that field carries that meaning.