[osis-users] Variant Readings

[David Troidl]

This has been an exciting and busy time at Open Scriptures. I'd like to 
try to keep the OSIS users up to date with what we are doing, as time 
permits. Currently, I am marking up the Ketiv/Qere readings in the 
Westminster Leningrad Code. Here is an example of the markup we were 
planning to use:

Gen.8.17 (excerpt)
<w lemma="3318">הוצא</w>
<note type="variant"><catchWord>הוצא</catchWord><rdg type="x-qere"><w 
lemma="3318">הַיְצֵ֣א</w></rdg></note>

I am wondering, though, about the maintenance of the catchWord. 
Repetition is often the enemy of maintenance. In updating, they could 
possibly fall out of sync. To the human reader, a slight discrepancy in 
a footnote would not be disastrous, but for software displaying variant 
readings, it could pose a problem. There are applications that present 
options to display the Primary reading, Secondary reading or All 
readings. These could be affected.

Another option we discussed is enclosing the K/Q pair in a container:

Judg.16.25 (excerpt)
<seg>
<w lemma="3588">כי</w>
<w lemma="2896">טוב</w>
<note type="variant"><rdg type="x-qere"><w 
lemma="2896">כְּ/ט֣וֹב</w></rdg></note>
</seg>

Here there are two ketivs combined in the qere. The grouping does keep 
them together, avoids the repetition, and would probably work, if the 
software were programmed around this convention. Still, the treatment of 
variant readings and notes is usually distinct.

I'm thinking of something like the following, possibly for the next 
release of OSIS:

Judg.16.25 (excerpt)
<seg type="ketiv" ID="a">
<w lemma="3588">כי</w>
<w lemma="2896">טוב</w>
</seg>
<rdg type="qere" idRef="a"><w lemma="2896">כְּ/ט֣וֹב</w></rdg>

This has the advantage of distinguishing the portion of the text 
referred to, creating a connection to the variant that can be validated, 
and promoting the rdg element to a status independent from the note. 
ID's are required to be unique for XML validation and IDREF's must refer 
to an ID that exists in the document. While Qere readings may possibly 
be thought of as notes that the Masoretes applied to the consonantal 
text, in the current practice of filling out the qeres and transferring 
the vowel points from the ketiv to the qere, they are becoming more of 
an integral part of the document.

This approach would also apply to other texts, such as the Tregelles 
Greek New Testament or the Robinson/Pierpont Byzantine Textform, where 
variant readings are commonly included. In those cases, the rdg could 
contain <rdg src="Steph" idRef="a">, to refer to the short name of a 
work specified in the header.

My questions are:
1) Which of the first two options would be best for the present?
2) Does the proposal make sense as an extension to OSIS?

Peace,

David