[Data-modeling] how much work done on modeling of personal names -- even for surname + given name?

[Christopher R. Maden]

Doing this “right” will involve some really heavy decomposition of the 
names.

I think there is some merit in just having a simple family name property 
on Person; in the case of Mrs. Mary Jones, née Smith, she would have 
links to both Smith and Jones.  (Strictly, Mrs. Mary Jones is a widow, 
as well...)

Otherwise, I think that the structured name needs to be a CVT.  The 
structured name CVT would have one property, name part, which is an 
ordered list of name parts.  Those parts would be co-typed as given 
name, maiden name, nickname, title, postnomial, etc.  The CVT could also 
have start and end dates:

title Miss, given Mary, given Ann, family Smith, 1963–1988
title Ms., given Mary, maiden Smith, family Jones, 1988–
title Mrs., given Jonathan, family Jones, 1988–

for someone *really* traditional...  or

title Ms., given Mary, given Ann, family Smith, 1963–1988
title Ms., given Mary, family Smith-Jones, 1988–1990
title Dr., given Mary, family Smith-Jones, postnomial Ph.D., 1990–1998
title Dr., given Mary, given Ann, family Smith, postnomial Ph.D., 1998–

~Chris
-- 
Christopher R. Maden
Data Architect
Freebase.com: <URL: http://www.freebase.com/ >
Metaweb Technologies, Inc. <URL: http://www.metaweb.com/ >