A listPerson is just a kind of list that happens to contain only person
elements. As such, it can appear in almost any place a list can appear,
including for example paragraphs, but it has never been allowed as a
direct child of teiHeader, nor is it now. If it appears in the header,
it's good practice to put it inside an element which supplies some
additional semantics about the role or function of the people listed
with reference to the document being described. So, if you put it inside
the <sourceDesc>, you are asserting that the people listed are the
source of your document, which for me makes sense only if you are
transcribing their conversation. As you suggest, a more plausible
place to put it is inside <profiledesc> if you are using it to supply
information about people mentioned in the document, or otherwise
participating in it in some sense. If you do that, though, you need to
wrap the listPerson inside a <particDesc> to separate it clearly from
the other children of <profileDesc>. And to answer your first question
(a) -- this is a general rule applying to quite a few other header
elements. You can either supply a semantically precise element such as
<listPerson>, or you can supply a generic prose description as one or
more <p> elements. But you cannot mix them. (Though you can of course
put your <listPerson> inside a <p>!)
On 13/06/16 10:36, Andreas Wagner wrote:
> Dear list,
> today I encountered validation complaints for a file that up to now
> has not been problematic:
> In this file with prosopographical information, I have a
> teiHeader/fileDesc/sourceDesc which contains only a single p element
> saying something along the lines of "born digital". Also, as an
> immediate child of teiHeader, I have a bunch of information inside a
> listPerson element.
> Apparently validation now only succeeds if I use listPerson to replace
> the p element in sourceDesc. But I am not sure I understand the motive
> for this:
> a) I am under the impression that an information like "born digital"
> is not well placed in any of the other available fields inside
> sourceDesc, but a plain <p> is not allowed if I also have a listPerson
> element in there.
> b) Why is listPerson meant to reside inside sourceDesc in the first
> place and not, for instance, in profileDesc?
> I am validating against tei-all.rng from the TEI framework extension
> in oXygen (http://www.tei-c.org/release/oxygen/updateSite.oxygen),
> which says:
> Schema generated from ODD source 2016-03-30T02:42:21Z. .
> TEI Edition: Version 3.0.0. Last updated on
> 29th March 2016, revision 89ba24e
> TEI Edition Location: http://www.tei-c.org/Vault/P5/Version 3.0.0/
> And the guidelines seem to indicate the same situation.
> Thanks for any insights.