[osis-core] Best practice to encode multiple endings to Mark, Rev, Ester, and Ps.151

[Chris Little]

I think the issue of multiple endings in Mark is identical to the issue of 
two Esthers in the NRSV, two books of Psalms in the Vulgate, and two each 
of Joshua, Judges, Tobit, & Daniel in Rahlfs' LXX.  Each of these is a 
work that contains multiple versions of the same book (in the case of 
Mark, it's just a part of the book, but still...).

Since it's a single work, the work portion of the osisID probably should 
not change.  But I think we also want to keep the verse reference portion 
of the osisID the same.  The best solution I could think of was to precede 
the portion of each ID that follows the work with another element to 
indicate membership within a collection.  So, for example, if I take my 
Rahlf's LXX and decide that recension B & Theodotian's translations are 
will be marked with B and all other books will be marked with A, I might 
have books like:

A.Gen
A.Exod
...
A.Josh
B.Josh
...
A.Dan
B.Dan

So when I run into B.Dan.1.1, I know it's Dan 1:1 in Theodotian's 
translation.

I know that I agreed a while back that the difference should be indicated 
in the work portion of the ID, but I've changed my mind since they are 
still part of the same work.

I think it would still be good, as you suggested, to indicate that they
are alternate endings.  You might surround each of them with a <div> or a
<seg> and indicate some type.  The "alternate-A", "alternate-B", etc.  
strategy might be better since the scheme is more systematic that
primary/secondary for people writing software and will provide more
possible alternates.  "alternate-1/2/3/..." would work too, and using
"variant" might make more sense to some people, though I've never seen it
applied to such large segments as the endings of Mark.  (I don't know
about marking this on "subType" if there is no "type" attribute since it
seems strange to have the former without the latter.)  I would probably
only mark this on the enclosing <div>, rather than elements it contains.

As for whether to mark it on only one or on both, I think it depends on 
whether the text itself exhibits favor towards one version.  If it does, 
then that version could be considered primary and wouldn't necessarily 
need marking.  My NRSV displays no preference, but my NASB seems to prefer 
the longer ending--so it varies.

> In Rev there is a single verse that is marked as an alternant but it is
> really just an additionally verse and not an alternant.  Should we use
> the same sort of strategy here?  If so should the "subtype" attribute
> value be "alternant", "alternant-additional", ...?

I've seen these single-verse alternates included only in footnotes, 
identified as variants only in footnotes, and identified (such as by 
enclosing in []'s) as a less-preferred reading.  Only in the last of these 
cases would I mark the verse as an alternate text, but I would try to be 
consistent with whatever practice is decided for the ending of Mark & 
such.
 
> Ps.151 although usually kept separate from Ps.1-150, seems to really be
> an alternant ending to Ps.  Thoughts?

I don't think this is ever the case.  Bibles I've seen either identify a 
151st Psalm within the book of Psalms (Ps.151) or contain a separate book 
"Psalm 151" (in which case it is Ps151.151).

> Lastly, Greek Esther also has the flavor of an alternant ending to
> Ester.
>
> How much of the separation of Greek Esther from Esther and Ps.151 from
> Ps.1-Ps.150 is a function of how they have traditionally been render vs
> the fact that they are truly separate "works" that happen to have the
> same name. Naturally it is possible create a "Protestant" version that
> did not have these portions to make the document more comfortable to
> groups that do not believe they are canonical.

I would also say that there is only one "Esther"; it just comes in two
widely varying versions.  If a work (e.g. the Douay, Vulgate, LXX, NJB,
etc.) calls a book Esther and it includes all 16 chapters of the Greek
version, it should include ids Esth.1-Esth.16 and have no indication that
certain portions are considered apocryphal by certain groups.  If a work
(e.g. the BHS, NIV, NASB, etc.) calls a book Esther and it include the 10
chapters of the Hebrew version, it should include ids Esth.1-Esth.10 and
have no indication that there are additional portions not included, which
are part of another document tradition.  Works like the KJV that have a
specific AddEsth section should have ids numbered according to whatever
system they employ internally (usually AddEsth.10.4-AddEsth.16 or 
AddEsth.11.2-AddEsth.11.1, depending on how you count).  If a
work includes both Hebrew & Greek Esther (as in the NRSV), I believe both
should be labeled "Esth" but preceded by another element, as I proposed
above, such as "he.Esth"/"grc.Esth".  In this case I'm not sure whether it 
would be appropriate to mark them as alternates, as might be done with the 
endings of Mark.  I probably wouldn't, when marking works like the NRSV, 
LXX, or Vulgate, but I can't really give a good defense of that position 
aside from not wanting to mark an entire book as "alternate".

--Chris