You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
{{ message }}
Repository navigation
Should Ramona Navarro vda. de Villegas keep vda. de Villegas in the family name? #565
"Vda." (viuda de, "widow of") is common in Spanish and Philippine records: Ramona Navarro vda. de Villegas is Ramona Navarro, widow of a Villegas. The no-comma form splits the compound, while the comma form already keeps it whole:
>>>parse("Ramona Navarro vda. de Villegas")
ParsedName(given='Ramona', middle='Navarro vda.', family='de Villegas')
>>>parse("Navarro vda. de Villegas, Ramona")
ParsedName(given='Ramona', family='Navarro vda. de Villegas')
So the paths disagree, the same pattern as #429, #430 and #432. Spanish y already gives the closest precedent: Juan García y Pérez reads family García y Pérez.
vda should not become a suffix word. de Villegas follows it, so a suffix reading would strand the husband's surname.
Options
(a) A connective, like y:vda. joins its neighbors, so family = Navarro vda. de Villegas. This matches the comma form and keeps the string intact.
(b) A marker clause, like née:de Villegas goes to family, and Navarro goes to maiden (or stays as middle). This is more interpretive: Philippine usage often treats the whole compound as the surname in use.
Open questions
Which spellings are listed: vda, vda., viuda? Vda capitalized mid-name is common.
Should vda. require a particle after it (de) before it joins?
What happens with no family before it (Ramona vda. de Villegas)?
Does Vda collide with any borne name? (Check under decisions.md#vocabulary-collisions.)
Activity