On Aug 2, 2005, at 10:22 AM, Matthew Cockerill wrote:
> (1) middle names (?)
>
> From: http://rdfweb.org/topic/NamesInFoaf it looks like the idea is
> that you provide a sequence of given names, the first being the
> traditional "first" name.
>
> <foaf:givenname>Jean-Claude</foaf:givenname><foaf:givenname>Pierre</
> foaf:givenname><foaf:givenname>Jacques</foaf:givenname>
> The fact that the first name is not special in any way could make
> processing more of a pain than necessary though.
> And then there's the question of which of the given names that people
> use as there preferred name.
> e.g. R Timothy Hunt, where Timothy is the middle name, but is used in
> preference to the first name.
>
> I guess what might b e needed there is an optional?) attribute of
> <foaf:givenname> to flag it as the one that is used by the person
> concerned (in some cases, several of the given names may be used I
> guess, so several could have this attribute..
Yes, you outline the problem well. "Middle name" isn't general enough
to cover even many Western names (J. Edgar Hoover, F. Scott Fitzgerald,
etc.).
There was an undergrad thesis out of the UK (by Richard Lennox) that I
gave some advice on where he settled on a very complex naming schema
based in part on the FOAF proposal and the discussion in the FRBR. For
this issue, he had a secondaryGivenname property IIRC.
> (2) suffixes (Jr, III)
Yup.
> (3) the fact that both the given name (in the case of reference
> lists), and the middle names (in most cases) will only be available in
> the source data as initials.
In my case, I actually manually track down full names for my authors if
possible. I'm not in the sciences, though, so the articles I cite are
typically single-authored.
> Assuming that all upper case letters are initials lets you get away
> without having a separate tag for initials compared to names, but
> would be ambiguous in any culture that has single letter names.
> Might be better to express clearly whether something is being asserted
> to be a name, or one or more initials.
Yes also. In citeproc (my XSLT-based citation processing system), I
make the pragmatic assumption that a given name structure with a single
character is an initial. It works, but is less-than-ideal.
Bruce
Received on Tue Aug 02 2005 - 14:28:42 EDT