Re: <p type="I_wish">

On 2016-10-08 09:26 AM, Lou Burnard wrote:
> Wouldn't @ana suffice for these?

Indeed. I would argue that in a case where "interior monologue is 
represented by italicized paragraphs", and the fact that it's interior 
monologue is a conclusion you draw from the font style, @ana is 
precisely the right thing.

This:

<p ana="#interiorMonologue">

is no harder or longer than this:

<p type="interiorMonologue">

and it has the slight advantage that while the documentation of your 
@type value, if it exists, will typically be in your schema, the 
documentation @ana points at may be in the header of your file. You can 
also use multiple pointers in @ana, allowing a more flexible and layered 
approach than is possible with @type and @subtype.

This argument obviously applies to a lot of situations in which @type 
and @subtype are used, but I tend more and more to the opinion that we 
should be using more pointers like this and fewer enumerated (and often 
sparsely-documented) @type values.

Cheers,
Martin

>
> On 07/10/16 23:03, Sewell, David R. (drs2n) wrote:
>> I'll admit that in the current case, and probably in many of the
>> previous ones, the pargraph typology I wanted to capture has local
>> significance but doesn't represent a generalizable typology. So from
>> the point of view of document interchange, possibly forcing an encoder
>> to come up with a more semantically specific TEI construct is
>> justifiable.
>>
>> But aren't there some pretty obvious examples that jump to mind? For
>> example, in a novel where interior monologue is represented by
>> italicized paragraphs, rather than treating that purely as an instance
>> of rendering? So
>>
>> <p type="interior_monologue">
>>
>> (with a presentational attribute as well if desired)? I'm sure there
>> are other gold-standard TEI ways of doing this, maybe using @ana
>> pointing to an <interp> element explaining that these are interior
>> monologues, which would have the advantage of being quite explicit
>> about things. But there are zillion TEI-encoded documents out there
>> with things like <div type="notes"> where the meaning of "notes" is
>> implicit in the @type value.
>>
>> OK, here's another example from something I worked on recently, a
>> volume of the press conferences of Woodrow Wilson. The volume uses
>> consistent, distinct, typographical paragraphing styles for questions
>> and answers. The most natural tagging would simply have been
>>
>> <p type="question">..</p>
>> <p type="answer">..</p>
>>
>> which is really not semantically equivalent to the TEI-valid
>> alternative I converted things to
>>
>> <sp who="#Wilson"><p>...</p></sp>
>> <sp who="#Press"><p>...</p></sp>
>>
>> because one can't be 100% sure that the speakers in each case are
>> "Wilson" and "the Press" -- the typography only lets you be sure that
>> one block of text is a question, and another is an answer. (I could
>> have gone with <ab> of course.)
>>
>> David
>>
>> On Fri, 7 Oct 2016, Martin Holmes wrote:
>>
>>> Quoting James, from an old closed ticket:
>>>
>>> [quote]
>>> Our default question when someone asks for att.typed membership for
>>> an element is to ask two questions:
>>> 1) Is the element repeatable?
>>> 2) Can we think of some reasonable examples of how someone might want
>>> to classify this element (both generally (@type) and specifically
>>> (@subtype)
>>> [/quote]
>>>
>>> <p> is clearly repeatable, so the second question is the one to focus
>>> on. If you have a couple of examples of paragraph typologies, please
>>> raise a ticket for this and Council will discuss it. It might be
>>> worth looking through some old related tickets to see the sorts of
>>> arguments that are likely to come up:
>>>
>>> <https://github.com/TEIC/TEI/issues?utf8=%E2%9C%93&q=is%3Aissue%20is%3Aclosed%20in%3Atitle%20type%20>
>>>
>>>
>>> Cheers,
>>> Martin
>>>
>>> On 2016-10-07 01:13 PM, Robinson, Peter wrote:
>>>> I am at one with David on this rant. It would be so cool to have
>>>> type as
>>>> a universal attribute, along with n/rend/xml:id and a few others. It
>>>> does rather surprise me to see how many universal attributes there are
>>>> (see
>>>> http://www.tei-c.org/release/doc/tei-p5-doc/en/html/ref-att.global.html)
>>>>
>>>> — and yet, @type is not one of them. Yes, I do recall the legions of
>>>> discussion on this point oh so many years ago. I was frustrated then, I
>>>> am frustrated now.
>>>>
>>>> Peter (frustratedly)
>>>>
>>>>> On Oct 7, 2016, at 9:56 PM, David Sewell <[email protected]
>>>>> <mailto:[email protected]>> wrote:
>>>>>
>>>>> Just venting a bit at the end of a longish week... For the X'th time
>>>>> in a decade or so, where X is a moderately large number, I find myself
>>>>> thinking "I do wish that TEI allowed @type on paragraphs, it would
>>>>> make my life so much easier!"
>>>>>
>>>>> Yes, I realize it's not going to happen, yes I'm going to use <ab
>>>>> type="whatever"> instead. It's a context where semantically the things
>>>>> really *are* typed paragraphs, not logical divs, not anonymous blocks.
>>>>>
>>>>> --
>>>>> David Sewell
>>>>> Manager of Digital Initiatives
>>>>> The University of Virginia Press
>>>>> Email: [email protected] <mailto:[email protected]>   Tel: +1
>>>>> 434 924 9973
>>>>> Web: http://www.upress.virginia.edu/rotunda
>>>>
>>>
>>