vd carries two real readings: the Dutch abbreviation of van der, and the British Volunteer Officers' Decoration. Trailing position always resolves to the decoration:
Berg, Jan vd given='Jan' family='Berg' suffix='vd'
Berg vd given='Berg' suffix='vd'
Jan vd Berg given='Jan' family='vd Berg' <- leading position is fine
Per #379, Berg, Jan vd should give family='vd Berg' — it is the same Dutch sorted form as Beethoven, Ludwig van. But the decoration reading has to keep working:
John Smith VD suffix='VD'
Col. John Smith VD suffix='VD'
The entry is legitimate
vd arrived in 962f235 ("add post-nominals from danhartropp in #7") as part of a UK honours list, alongside RD, VRD, TD and ED — the reserve/volunteer decoration family. #7 was scraping the UK parliament website. This is not a bulk-import artifact like rai in #342, and it should not simply be removed.
Neither existing mechanism arbitrates it
The positional rule described in #289 — an ambiguous acronym is a credential when removing it still leaves a given and a family name — fires correctly by its own logic here, since removing vd leaves Jan + Berg. It has no way to know about the Dutch convention.
The period form does not separate them either: both vd and V.D. are attested spellings, on both sides of the ambiguity.
A signal that might
Dutch tussenvoegsels are lowercase (vd, v.d.) while the decoration is uppercase (VD). That is exactly the letter-case signal #289 proposes, so this may be a case for that mechanism rather than a new one. Note #289's current scope is the four words in suffix_acronyms_ambiguous (do, ed, jd, ma); vd is not among them, so adopting it there would widen that issue.
Not the same as mc
mc looks like the same shape and is not. Moving it to the never-given half per #360 changes no suffix reading — measured:
Mc Donald shipped: given='Mc' family='Donald' mc moved: family='Mc Donald' (the #360 fix)
John Smith MC shipped: suffix='MC' mc moved: suffix='MC'
Donald mc shipped: suffix='mc' mc moved: suffix='mc'
Trailing mc routes on suffix_acronyms membership, which curation does not touch. There is also no naming convention that writes Mc Donald as Donald mc, so the collision is theoretical there and live here.
Related
vdcarries two real readings: the Dutch abbreviation ofvan der, and the British Volunteer Officers' Decoration. Trailing position always resolves to the decoration:Per #379,
Berg, Jan vdshould givefamily='vd Berg'— it is the same Dutch sorted form asBeethoven, Ludwig van. But the decoration reading has to keep working:The entry is legitimate
vdarrived in962f235("add post-nominals from danhartropp in #7") as part of a UK honours list, alongsideRD,VRD,TDandED— the reserve/volunteer decoration family. #7 was scraping the UK parliament website. This is not a bulk-import artifact likeraiin #342, and it should not simply be removed.Neither existing mechanism arbitrates it
The positional rule described in #289 — an ambiguous acronym is a credential when removing it still leaves a given and a family name — fires correctly by its own logic here, since removing
vdleavesJan+Berg. It has no way to know about the Dutch convention.The period form does not separate them either: both
vdandV.D.are attested spellings, on both sides of the ambiguity.A signal that might
Dutch tussenvoegsels are lowercase (
vd,v.d.) while the decoration is uppercase (VD). That is exactly the letter-case signal #289 proposes, so this may be a case for that mechanism rather than a new one. Note #289's current scope is the four words insuffix_acronyms_ambiguous(do,ed,jd,ma);vdis not among them, so adopting it there would widen that issue.Not the same as
mcmclooks like the same shape and is not. Moving it to the never-given half per #360 changes no suffix reading — measured:Trailing
mcroutes onsuffix_acronymsmembership, which curation does not touch. There is also no naming convention that writesMc DonaldasDonald mc, so the collision is theoretical there and live here.Related
Beethoven, Ludwig vanputs the tussenvoegsel inmiddleinstead of the family name #379 — the tussenvoegsel comma form, which deferred this case explicitlyRaiis parsed as a post-nominal suffix, consuming a common South Asian surname #342 — contrast: a genuinely bogus suffix entryMc DonaldandSte Marieread the particle as the given name #360 — the particle curation that raised the question