Open martindholmes opened 3 years ago
Leaving this unassigned for now, with intent to assign to either @bleekere or @sabineseifert in Jan 22.
To just do what this ticket asks (add <person>
to att.typed) is easy. But there should probably be discussion of
<listPerson>
if not present@type
of parent <listPerson>
?@type
and @role
plus an example or two.
adapted from https://tei-c.org/release/doc/tei-p5-doc/en/html/ND.html#NDPERSE as I was just trying to do just that and realized no person@type
<person xml:id="person_FAS" type="fictional" role="attorney">
<persName>Adam Schiff</persName>
<note>District Attorney for <placeName>Manhattan</placeName> in
seasons 1 to 10 of <title>Law and Order</title>.</note>
</person>
Here is an example with @type
and @role
on <person>
to showcase and discuss possible usage and differences:
<listPerson type="fictional">
<person type="main-act" role="actress waitress sales-representative">
<persName>Penny Hofstadter</persName>
</person>
<person type="main-act" role="theoretical-physicist">
<persName>Sheldon Cooper</persName>
</person>
<person type="supporting-act" role="human-resources-representative">
<persName>Janine Davis</persName>
</person>
<person type="supporting-act" role="plasma-physicist">
<persName>Barry Kripke</persName>
</person>
</listPerson>
<listPerson type="real">
<person type="guest-appearance" role="theoretical-physicist cosmologist">
<persName>Stephen Hawking</persName>
</person>
<person type="guest-appearance" role="actress">
<persName>Carrie Fisher</persName>
</person>
</listPerson>
Responding to @sydb's questions, @peterstadler and I propose:
@type
on <person>
should be understood as subcategory / as complementing the categorization of @type
on <listPerson>
; it inherits the semantics of the parent's statement and does not override and must not be in contradiction@type
and @role
: two possible types of classifications, @type
being the more generic one and @role
the more specific / targeting one
Just like
<org>
and<place>
, person should be a member of att.typed; this would enable type to be distinguished from role, and also enable us to avoid the tendency to add@type
and@subtype
to name or persName elements to get around the absence of it. Although<listPerson>
can be used to organize people by type, that is ungainly when you have many combinations of@type
and@subtype
.