Some thoughts on an excellent question follow. They represent musings
more than anything else.
Probably best if resp= of <note> points to the xml:id= of either:
* a <respStmt> or its syntactic sugar <author>, <editor>, etc.
(typically, but not necessarily in the <teiHeader> of the same
* a <person> (typically, but not necessarily, in a <listPerson> that
lists people involved with the project)
Some will argue that, because the description says "... a person ..."
that it should point to the <name> or <persName> inside <respStmt>,
<author>, or <editor>, rather than to the <respStmt> itself.
Furthermore, such listeralists will say that it should not point to
But I will argue that it is the description of the values of resp= of
<note> that is at fault: it is both a bit too precise (saying
"person") and not sufficiently precise (which element?) at the same
A single individual may well perform multiple roles in a document's
production, and it may be appropriate to record which hat was being
worn when a given <note> was produced. (And it may not -- I am not
claiming it is inappropriate to point to a <persName> inside a
<respStmt> or <author>, only that it is appropriate to point a
<respStmt> or <author>.)
There are also circumstances where it makes sense to attribute
responsibility to an entire organization (e.g., the editorial team or
web-hosting institution), in which case it may make sense to point to
a <respStmt> that contains an <orgName> rather than a <persName>.
I will try to post an example shortly.