[sword-devel] Module TischMorph & other Ancient Greek modules - getting rid of the quotation mark?

[pierre amadio]

Hi there.

It looks like an old and complex issue:
https://corp.unicode.org/pipermail/unicode/2019-January/007563.html

I had a quick look at the thread and still don't understand what
should be the correct character to use :-)

If I understand correctly, there is no such thing as quotation mark
character in ancient Greek (*)  , so there should be no problem
confusing an elision marker with an end of quote mark anyway.

(*) There can be some sort of quotation mark such as here, in the
margin of a codex vaticanus page, on the middle part of the left
column (luke 3 citing the book of Isaiah)
https://digi.vatlib.it/view/MSS_Vat.gr.1209/1313

On Sun, 16 Mar 2025 at 09:08, David Haslam <dfhdfh at protonmail.com> wrote:
>
> Indeed, getting all the upstream Greek NT sources fixed would be ideal, but this "category mistake" has become so entrenched since before the advent of computers, I suspect we'd have an uphill struggle.
>
> It's all the more frustrating when one realizes that there are only 21 distinct words in a Greek NT where this inappropriate Unicode character was used! That these fairly short words are quite common is what accounts for the overall total of 1219.
>
> That figure was calculated after I used BabelPad Tools to perform a Word Frequency analysis both before and after replacing each U+2019 by U+0315 as part of my experiment.
>
> To illustrate this, I have attached the resulting diff file.
>
> NB. Because U+2019 is not a "word character", it's not present in the before analysis.
>
> Refer also to the transcript of the second part of my AI chat with Leo, as per my email sent late last night.
>
> How can we persuade the wider community of NT Scholarship and Textual Criticism to fix this long-standing category mistake?
>
> Best regards,
>
> David
>
> Sent with Proton Mail secure email.
>
> On Sunday, March 16th, 2025 at 12:26 AM, Greg Hellings <greg.hellings at gmail.com> wrote:
>
> It sounds like the text is using the wrong character. Would it not be better to go back to the source text and fix it there? That would benefit everyone who consumes that text, and not just people looking at the appearance of the text through a Sword client with sufficiently up-to-date filters.
>
> Why is the text using an apostrophe character in the first place, rather than the appropriate accent character?
>
> --Greg
>
> On Sat, Mar 15, 2025 at 1:46 PM David Haslam <dfhdfh at protonmail.com> wrote:
>>
>> Further to my message sent to the modules team, I have suggestion for SWORD developers.
>>
>> Modify how GlobalOptionFilter=UTF8GreekAccents works with respect to U+2019.
>>
>> Currently the filter doesn't care how this codepoint is being used semantically.
>>
>> We could make a practical change as follows:
>>
>> Rule: Do not hide U+2019 if the preceding character is not a Greek letter.
>>
>> However, that would not be sufficient to cover all the cases in the StatResGNT module currently in CrossWire Beta.
>>
>> Here's one that's better:
>>
>> Rule: Do not hide U+2019 if the preceding character is not a word character.
>>
>> With this in operation, the source text would not need to be changed in those places where the RIGHT SINGLE QUOTATION MARK is immediately after a punctuation mark.
>>
>> Then what about the remaining places where it is immediately after a Greek letter?
>>
>> A simple kludge for an improved source text would be to insert a zero width space or a zero width non joiner. Doing so would protect the quotation mark from being treated contrarily as a Greek diacritic when its purpose is simply to close a quotation.
>>
>>
>> Best regards,
>>
>> David
>>
>> Sent with Proton Mail secure email.
>>
>> On Saturday, March 15th, 2025 at 6:02 PM, David Haslam <dfhdfh at protonmail.com> wrote:
>>
>> Thinking about the teleological conflict caused by using U+2019 RIGHT SINGLE QUOTATION MARK as a Greek Diacritic,
>> I wish to suggest that in modules where this was the case, that we might replace all U+2019 by U+0315 COMBINING COMMA ABOVE RIGHT.
>>
>> This displays very similarly, viz
>>
>> U+2019 παρ’
>> U+0315 παρ̕
>>
>> NB. The latter was not already used anywhere in the TischMorph module.
>>
>> The advantage would be that if we did this to all the Ancient Greek modules and to any other module that makes the same contrary use of U+2019,
>> then eventually, we might update the SWORD engine such that it no longer hides it with GlobalOptionFilter=UTF8GreekAccents.
>>
>> This would free up U+2019 for it's proper use as a RIGHT SINGLE QUOTATION MARK to end speech begun with U+2018 LEFT SINGLE QUOTATION MARK.
>> We have just encountered this use in Beta module StatResGNT which contains sections of text marked with both DOUBLE and SINGLE QUOTATION marks.
>>
>> I propose therefore that we perform an experiment, by implementing my suggestion first in module TischMorph, last updated 2012-01-09.
>> This has 1218 instances of U+2019 and in each case, it's used contrarily as a Greek Diacritic.
>>
>> Here's a counted list of the Greek letters that are followed by a U+2019.
>>
>> 0031 δ’
>> 0109 θ’
>> 0232 τ’
>> 0287 π’
>> 0221 λ’
>> 0135 φ’
>> 0059 ρ’
>> 0144 ι’
>>
>>
>> This is how they would look if the proposed replacements were made:
>>
>> 0031 δ̕
>> 0109 θ̕
>> 0232 τ̕
>> 0287 π̕
>> 0221 λ̕
>> 0135 φ̕
>> 0059 ρ̕
>> 0144 ι̕
>>
>>
>> NB. The upstream source for this module is no longer found at http://morphgnt.org/
>> The URL now just displays a 404 message saying "There isn't a GitHub Pages site here."
>> If we can no longer find where it may have moved, we might have to use a "round trip" of mod2imp and imp2vs to rebuild the module with my suggested change.
>>
>> If my suggestion has any merit, we should also first discuss it with NT scholars such as our friends at STEP Bible and their friends at Tyndale House.
>>
>> Best regards,
>>
>> David
>>
>> Sent with Proton Mail secure email.
>>
>>
>> _______________________________________________
>> sword-devel mailing list: sword-devel at crosswire.org
>> http://crosswire.org/mailman/listinfo/sword-devel
>> Instructions to unsubscribe/change your settings at above page
>
>
> _______________________________________________
> sword-devel mailing list: sword-devel at crosswire.org
> http://crosswire.org/mailman/listinfo/sword-devel
> Instructions to unsubscribe/change your settings at above page