[tei-council] on regularizing names

[Christian Wittern]

Syd Bauman <Syd_Bauman at Brown.edu> writes:

>> Could another option be a twist on (j)? Instead of a pointer to a
>> regularization, with no key= on the <regName>, what about a pointer
>> to a regularization, with no key= on the <persName>? It seems
>> having key= on both <persName> and <regName> is redundant, and
>> having to include a key= and reg= on every <persName> is excessive
>> when one could include just the reg= and find the key= in the
>> referenced <regName>.
>
> Yes, for a project that is both keying and regularizing it would make
> good sense to use reg= on <persName> (w/o key=) and key= on
> <regName>. But we need to permit key= on <persName> for those
> projects who are keying, and either not regularizing or are
> performing regularization via whatever database the key= points to.

Sorry guys, I do not get you here.  

To me @key on <regName> seems outright wrong.  @key is supposed to
point to records of persons, whereas <regName> has only to do with
the *regularization* of names.  So I do not understand what role @key
has on <regName>, other than a convenient shortcut for those special
cases where only one person of that specific name exist in your
records (but you are out of luck if a second one shows up).  If that
is indeed the use case, I do not think we should recommend this.

Christian


-- 

 Christian Wittern 
 Institute for Research in Humanities, Kyoto University
 47 Higashiogura-cho, Kitashirakawa, Sakyo-ku, Kyoto 606-8265, JAPAN