[osis-core] paragraph break defended once again.

[Chris Little]

Troy,

Sorry, been too busy & ill-connected to get a reply out.

Troy A. Griffitts wrote:
>> But what's more, I really can't conceive of any use for it.  Could 
>> someone think up a use case for it?  I would concede if there were 
>> some actual use in marking e.g. manuscripts.
> 
> http://www-user.uni-bremen.de/~wie/Vaticanus/general.html#div

I see some different sorts of text divisions described.  The description 
is too limited to know for certain, but I would assume these divisions 
can still be placed inside a container.  If they are meaningful units, 
by any criterion, they probably have a beginning and an end.

In any case, they are not paragraphs and a paragraph break marker is 
inappropriate.

>>  There is no ambiguity as to what is intended by their use of their 
>> DTD in their document--in fact it's probably the cleanest & most 
>> lucidly marked document I've ever encountered.
> 
> Well, to cite an example for you.  Just looking briefly, reading the 
> text would make one assume that there should be a paragraph break 
> between Mat. 1 and Mat 2.  Also, if I had to guess, I would probably not 
> place a paragraph break between Mat. 8 and 9.
> 
> I just don't want to have to read the text and guess-- that's all.

In the specific case for which you want this marker, the NASB, there's 
no ambiguity though.  Chapters start new paragraphs and new paragraphs 
are started at each paragraph break marker according to the Lockman 
markup.  Ask someone at Lockman and I'm confident they will tell you the 
same.  They make the assumption that chapter divisions are meaningful units.

In the case of NASB, paragraphs span neither the Matt.1/Matt.2 nor 
Matt.8/Matt.9 chapter boundaries.  In the case of the first, there is a 
paragraph spanning Matt.1.18-Matt.1.25 and another from 
Matt.2.1-Matt.2.6.  In the case of the latter, there is a paragraph 
spanning Matt.8.28-Matt.8.34 and another from Matt.9.1-Matt.9.8.

Rendering in print versions and in the version for Logos (which is 
actually published by Lockman, IIRC) reflect this layout.

Even in the case of the most controversial chapter division in the 
Bible, the Rev.12/Rev.13 division, which most people now think puts a 
verse from Rev.12 inside of Rev.13, grouped with what should have been 
the first verse in Rev.13.1, Lockman starts a new paragraph at the 
beginning of the chapter.  (They also end it after half a verse and 
start a new paragraph, but both paragraphs are in Rev.13 and in the same 
section.)

>> Let me put it this way: Paragraph breaks necessarily follow (or 
>> possibly come between?) paragraphs.  In order to have one, you must 
>> have the other.  Paragraphs are blocks of content that deserve to go 
>> inside a container (<p>).  What Troy is suggesting is that there is no 
>> such thing as a paragraph in the NASB--only paragraph breaks.
> 
> I am not suggesting that the NASB does not have a concept of a paragraph 
> container.  I'm suggesting that they do not have unambiguous markup of 
> these, and I don't feel it is the encoder's task to unambiguate the 
> containers.
> 
> I also see other uses for a paragraph break marker, like in mss markup 
> or other fragment markup (possibly including modern document segments)

Using <milestone type="paraBreak"/> for mss markup in texts like the one 
you cited seems like it requires an excess of interpretation.  They may 
act like paragraphs, but they don't seem to be called paragraphs and 
they don't look like paragraphs.  If we want a milestone marker for mss, 
I would concede to whatever is thought appropriate by people who know 
more about the subject than I.  I would be inclined to either punt to 
the mss markup WG--they seem like a lazy bunch since they never do any 
work, so maybe they just need encouragement :) --or to adopt a more 
generic section break marker.  I have a feeling that section break is 
too generic to be meaningful (absent myriad subTypes) to be useful for 
mss markup, while paragraph break is inappropriate on the basis of the 
sample given.

For modern document segments, I would suggest using milestones.  We 
shouldn't be converting </p> to <milestone type="paraBreak"/> but rather 
to <p sID="..."/>.

--Chris