Direct answer
A canonical name is the preferred primary name in a particular record and context. An alternative name is another current name for the same entity. A former name is no longer the current primary name and is preserved with a date range. A translated name is the form used in another language; it does not create a new entity.
In plain language
A person's official name, everyday short name, surname before marriage and form written in another language belong in different fields. If all are placed in one 'name' box, no one knows which to use, where or when.
Why this matters
Without name types, a former name may be treated as current, two entities may be merged incorrectly, or the language pages of one entity may appear disconnected.
Do not confuse
- A canonical name is not necessarily one universal name for every country and legal context.
- An alternative name may be current; a former name is historical.
- Translation and transliteration are different operations.
What should you do?
- Record each name's type, language and script.
- Add start and end dates to former names.
- Separate the legal name from the name displayed to users.
- Use valid BCP 47 forms for language tags.
- Link every name to the same stable entity identity.
How do you audit it?
- Are the name types explicit?
- Is a former name displayed as current?
- Are the language and script tags correct?
- Did a translation create a new entity record?
- Was the wrong name used in a legal context?
Limit
The choice of canonical name depends on context. Separate legal rules may apply to brands, company names and personal names.
Remember in one sentence
Every name may point to the same entity, but not every name serves the same purpose.
Sources for this record
- S08Schema.org, *Data Model*Standard and community vocabulary
- S12ORCID, *ORCID Trust*Identity infrastructure
- S13ROR, *ROR Data Structure*Identity infrastructure
- S14DataCite, *Versioning*Identity infrastructure
- S16IETF, RFC 5646, *Tags for Identifying Languages*Standard