[sword-devel] KJV 2006 5th beta

[DM Smith]

Chris,
    I really appreciate you input! I do have a few questions and some 
thoughts I'd like you to respond to. See below.
DM

Chris Little wrote:
> DM Smith wrote:
>> This beta corrects mistakes reported against the last beta. And also 
>> does the following:
>> Divine Name:
>> All divine names are tagged with <divineName> (most were already).
>> If the underlying strong's number is 3050 (YH), 3068 or 3069 (YHWH), 
>> I have added the type x-yhwh.
>> For those instances where the name is a compound with YHWH, the type 
>> is x-yhwh and the subtype is a transliteration of the hebrew of the 
>> compound part, e.g. x-tsidkenu.
>> There were also a few cases of JESUS in all caps. This is with the 
>> type x-jesus.
>> When the divine name  is in a <transChange> or a <note> I have tagged 
>> it without a type.
>> I also wrapped 2 instances of BRANCH in all caps with divineName as 
>> it appears that this is what the KJV was trying to communicate in 
>> making it all caps.
>> In all of these instances I have changed the word to begin with a 
>> capital letter and be followed with lower case. The intention is that 
>> these should be rendered with small-caps.
>> Questions: Should these be left as all caps? Can the front-ends 
>> render this?
>
> <divineName> is currently only intended for translations of Yahweh 
> (Strong's numbers 3050, 3068, & 3069). We debated whether it should be 
> used for marking Jesus or other deities named in the Bible or other 
> works, but ultimately decided that we should limit it to exclusively 
> Yahweh. We discussed adding types but, ultimately, decided not to 
> since the only valid type this leaves was "yhwh" or "Yahweh".
>
> If you think the situation should be otherwise, you should make the 
> case for a change on osis-user.

It's not that I think it should be otherwise. I was following the 
guidance of the OSIS user's manual. I just checked the most recent one 
which give the guidance
(And I quote):
> 13.4. divineName
> The divineName element is used only for the Deity. Angels, demons, 
> idols, and the like should be tagged
> with <name type='nonhuman'> For example:
> <divineName>El Shaddai</divineName>

In this case El Shaddai is not YHWH. See 
http://www.hebrew4christians.com/Names_of_G-d/El/el.html for the Hebrew.
Given "El Shaddai" there a lot more that could be tagged (which I did not.)

Also, in another place in the manual it explicitly give an example with 
a divineName having a type of x-yhwh. To me this implies that the tag is 
to for more than just the yhwh.

I don't have a problem with limiting it to YHWH. (It is an easy change.) 
But what about when it is given as the name of God in combination with 
another word as in yhwh-tsidkenu, The LORD is our righteousness. (i.e. 
strong's numbers 3070-3074)

As to Jesus, I only tagged the two occurrences where the KJV had it in 
all caps. In context, these are being ascribed the name of the Deity.

>> Inscriptions:
>> I have identified all inscriptions that were in all caps in the KJV. 
>> I have wrapped these with <inscription> and changed the all caps to 
>> Title Case. The intention is that these should be rendered with 
>> small-caps, perhaps emphasized with bold.
>> Question: Should these be left as all caps? Can the front-ends render 
>> this?
>
> From these and the later comments about these being all caps in the 
> printed KJV text, it sounds like the KJV2006 should have 
> <inscription>ALL CAPS</inscription>.

OK.

>
>> Paragraphs:
>> I have compared the paragraph marks against two other e-texts and 
>> then against a print Old Scofield. The result is:
>> All paragraph marks starting a chapter have been deleted.
>
> Why? I guess I don't remember whether you decided to encode paragraphs 
> (<p>...</p>) or mark pilcrows via milestones (<milestone 
> type="x-p"/>?). The former is valid OSIS. The latter is meaningless 
> for interchange. Interchange formats are pointless if we don't all 
> actually use the shared vocabulary.

I based the decision on a few things:
1) Your input.
2) Troy's input.
3) The printed text of the KJV.
4) A limitation of osis2mod (it has a bug where 
<p><verse>...</verse>....</p>, does not work. This has been reported for 
a while.)
5) Such a decision does not prevent using <p> as a container, later.

While it may be fair to infer that a paragraph begins each chapter and a 
chapter ends a paragraph, it is not clear from the text that this is the 
case.

I chose the milestone element based on the OSIS manual.

OSIS gives <milestone> as an element to preserve locations and markup of 
the source.
OSIS gives <milestone> the marker attribute to preserve the marker that 
is in the original text. (Though the only examples are for quotation marks.)
OSIS provides no particular type for a marker of the start of a 
paragraph and I'll ask for this to be discussed on the osis-user's group.

To quote the manual:
> The milestone element also has a marker attribute. This attribute is 
> used to record typographic features in a
> text that cannot be duplicated by automated scripting. If a text 
> inconsistently using a paragraph marker for
> example, it would not be possible to duplicate inconsistent usage 
> without that information being recorded as
> part of the encoding process.

The other way I could preserve the marker is to insert it literally into 
the text. As with <seg type="x-p">¶</seg> or just ¶.

I agree with you that all type extensions (e.g. x-yyy) are to be avoided 
if possible.

>
> This is probably another thing to bring up on osis-user if you thing a 
> change is needed.
>
>> Acrostic titles in Psalm 119:
>> I have added the attribute xlit to each <w> element surrounding the 
>> transliteration of the Hebrew alphabet letter (e.g. ALEPH) and set 
>> the value to the Hebrew character. This is not intended for display, 
>> but for completeness.
>
> Something doesn't seem quite right here. I would have no problem with 
> <w xlit="aleph">א</w>, but <w xlit="א">ALEPH</w> doesn't seem right. 
> And I don't really know that I have the right answer here, but 
> gloss="he:א" or lemma="he:א" seem like they would be slightly better. 
> Ultimately, I think the n attribute is the closest to the correct 
> place to hang the Hebrew letter (n="א"), but it may be better on the 
> container of the whole acrostic section.

I am not at all sure what's best here, either. This doesn't feel right 
to me either. And I can leave it out altogether for another day. It is 
my addition to the markup, as it is not in the KJV 1769 (or 1611 for 
that matter.)

I took my lead from some ThML I had seen which marked every 
transliteration into English letters with the language it was derived 
from, though not with the actual word it was from.

I took a look at the manual with regard to the n attribute. The guidance 
with regard to its use is vague. It is used as a note marker (which the 
manual discourages), as the original numbering of a text that has a 
different numbering, as the actual number of a chapter.

So, for now, I'll leave it out. I'll check with osis-users to see what 
guidance they provide.
>