This is similar to what SBLGNT did, and the computer end of this is straightforward.Eeli Kaikkonen wrote: ↑August 7th, 2018, 2:59 pmWe can re-interpret my question "why wouldn't that work today" to mean "if it produced a work with credibility then, why wouldn't it work in similar way today?"
That said, Alan Bunning's text is similar to the Nestle-Aland and is computed according to clearly identifiable rules. The computing end of this isn't the hard part. I assume that a good text-critical scholar is needed to (1) help identify the most relevant parameters and (2) apply human judgment at the end.
The NIV text differs in about 219 places. Alan Bunning's in about 500 places. That might be close enough.Eeli Kaikkonen wrote: ↑August 7th, 2018, 2:59 pm Just think about it - NA is kind of a consensus and compromise in itself. We can suppose that people want actual textual criticism and scholarly informed decisions behind a text, whether it's necessary for their purposes or not. Having several ones to choose from (as much as the licences permit) is on the other hand great, but it may make the situation more difficult because if you don't know textual criticism yourself you're unsure about the choice. What would be a better option than to combine more scholarly opinions? Psychologically thinking NA+SBL+TH could feel even more trustworthy than NA alone. The situation isn't the same with combined manuscripts, whatever the real merits would be. And I think it's important to get a text with as few differences with NA as possible while still keeping the work none-derivative.
I suspect it should at least be reviewed by someone most text-critical scholars would acccept as a text-critical scholar.
All three texts are available in digital form, the "for free" part may be an issue. I wouldn't be surprised if it were relatively easy to negotiate legal access for this purpose.Eeli Kaikkonen wrote: ↑August 7th, 2018, 2:59 pmIn practical level I might be interested in doing it. But are all the three texts available for free in a form which makes it possible to normalize them to a common lemma/parsing form to be compared? Plain text comparison is out of question because at least TH uses so much different spellings. Some manual work is of course always necessary.
I think the diffs will be available in TANTT. It's not all there yet.
TANTT - Tyndale Amalgamated NT Tagged texts
Greek text created from the SBLGNT+apparatus, following the decisions made by NA28, listing the major editions that also use that form (SBL, Treg, TR, Byz, WH, NA28). Variants are being added from major editions plus the 1st 4 centuries of MSS (from Bunning). All words are tagged lexically (extended Strong linked to LSJ) and morphologically (Robinson based on Tauber plus a few missing details) plus context-sensitive meanings for words with more than one meaning. An independant scholar checked the result against NA28 and pointed out a few differences which were fixed (see the issues in github).