Should a lone name word read as the given name under a family-first pack, so parser_for(locales.VI).parse("Anh") gives given Anh? #1111

A name of one word reads as the given name by default and as the family name under a family-first order: rules/O5, the lone name word's convention, which follows the declared order. No Policy field sets the convention apart from the order, so locales.VI, which declares FAMILY_FIRST_GIVEN_LAST, reads Anh as the family name, though a Vietnamese name written alone is usually the given name, the name a person is addressed by.

Each carries the given-or-family ambiguity. CLDR 48's vi names file states the one-word Anh and Sinbad given in three orders each. The pack loses 6 of the file's 24 renderings in the CLDR per-pack view (16 agree by default, 10 with the pack), all written given-first. 2 of them are this question's cost, Anh and Sinbad, which it reads family. The other 4 are multi-word names CLDR writes given-first (Käthe Müller, Anh Nguyễn), which the declared order reverses whatever a lone word reads as. The sorting and surname-first renderings of the two one-word names disagree with the pack and without it alike. (Corrected 2026-11-06 by #1108's design-docs review; this issue first said all 6 one-word renderings.) rules/O5 records the cost as an Accepted line (#1108).

Proposal, in rule shape

• When: a name has exactly one name word that nothing else decides, under a family-first order.
• Then: the word reads as the field a new Policy field names (a lone_word of "given" or "family"; unset, it follows the order, as today), which a [rules] table may set, so vi.toml sets "given".
• Because: the order a name is written in says nothing about which field a lone word fills; a culture can write names family-first and still write a person alone by the given name.
• Except: the cases rules/O5 already lists (a title, a script's own convention, a nickname, a maiden name, the word's own claim) still decide first.
• Accepted: a lone surname written alone (Nguyen) reads given under such a pack, reported as now.
• Open: does any other family-first convention (Chinese, Korean, Japanese, Hungarian in Latin letters) want the given name too, or is this Vietnamese only? Is a Policy field the right home, or should the convention follow the order everywhere as it does now? What do the Vietnamese local corpora (UIT-ViNames, the OpenSanctions deputies) hold for one-word names?
• Measure: the one-word names in the Vietnamese local corpora and in CLDR's vi file, read with and without the field.

Related: #1108, rules/O5.

Measured on #1108's branch at 7668791, 2026-11-05.