PATr variations

[doday]

I've been reading the Patriarchal Greek New Testament (PATr) that was published by the Patriarchal Press of the Ecumenical Patriarchate of Constantinople in 1904. My understanding is that it was based mainly on lectionary manuscripts from the ninth to the sixteenth centuries and that it was corrected in 1912 by Professor Vasileios Antoniades of the Theological School of Chalki. As such, it is often abbreviated PATr 1904/1912.

However, I've noticed there appear to be several variations / streams of this text. Sources I've identified, all of which claim to be PATr 1904 (with later corrections), include:

• eBible: "1904 Patriarchal Greek New Testament with 20 corrections from later editions." Editor unknown.
• GOA: Greek Orthodox Archdiocese of America (GOA) website. "The Greek New Testament displayed is the authorized 1904 text of the Ecumenical Patriarchate of Constantinople. The Patriarchal text has been made available courtesy of the Greek Bible society and was digitized in XML in cooperation with the Greek Orthodox Archdiocese Department of Internet Ministries, the Greek BIble Society and the American BIble Society IT Department's OSIS project. The Open Scriptural Information Standard (OSIS) was developed by the Bible Technologies Group in co-sponsorship with the American Bible Society and the Society of Biblical Literature."
• HBS: Hellenic Bible Society (HBS) website. Πατριαρχικό Κείμενο (Έκδοση Αντωνιάδη, 1904). "Copyrighted by the Hellenic Bible Society, 2017." This text appears to be getting fetched via API calls to the American Bible Society's Bible API.
• Logos: Logos edition of The Patriarchal Greek New Testament (PATr 1904/1912) by the Hellenic Bible Society and Logos Bible Software.

There are others, too, some of which almost certainly are not PATr, but for the purposes of this post I'll only use the above editions to distinguish each of the compared variations.

Here are just a few examples:

Matthew 22:32 —

eBible // GOA // Logos
... οὐκ ἔστιν ὁ Θεὸς Θεὸς νεκρῶν, ἀλλὰ ζώντων.

HBS
... οὐκ ἔστιν ὁ Θεὸς νεκρῶν, ἀλλὰ ζώντων.

Mark 4:3—

eBible // Logos
... ἐξῆλθεν ὁ σπείρων τοῦ σπεῖραι.

GOA // HBS
... ἐξῆθεν ὁ σπείρων τοῦ σπεῖραι.

John 19:31—

eBible // HBS // Logos
... ἡ ἡμέρα ἐκείνου τοῦ σαββάτου....

GOA
... ἡ ἡμέρα ἐκείνη τοῦ σαββάτου....

Acts 4:36—

eBible // HBS // Logos
... ἀπὸ τῶν ἀποστόλων....

GOA
... ὑπὸ τῶν ἀποστόλων....

Acts 16:34—

eBible // HBS // Logos
... καὶ ἠγαλλιᾶτο πανοικὶ πεπιστευκὼς τῷ Θεῷ.

GOA
... καὶ ἠγαλλιάσατο πανοικὶ πεπιστευκὼς τῷ Θεῷ.

Acts 26:30—

eBible // GOA
... Βερνίκη ....

Logos // HBS
... Βερενίκη ....

2 Corinthians 8:2–

eBible // HBS // Logos
... βάθους πτωχεία αὐτῶν....

GOA
... βάθος πτωχεία αὐτῶν....

Hebrews 8:11—

eBible // HBS // Logos
καὶ οὐ μὴ διδάξωσιν....

GOA
καὶ οὐ μὴ διδάξουσιν....

James 4:14—

eBible // GOA
... ἀτμὶς γάρ ἔσται ....

HBS // Logos
... ἀτμὶς γάρ ἐστιν....

I've only included a small number of examples and have excluded certain differences from consideration such as movable nu and other minor orthographical differences / elision (e.g., εἰμὴ vs. εἰ μὴ). It's perplexing that they don't all differ in a consistent manner. There appears to be four relatively distinct streams of PATr in the examples shown, and two of them are "official" sources (yet differ).

What's the deal? Anyone else notice this before? These all claim to be reproductions of the PATr 1904/1912, but clearly have more variations between them that don't support them all being the same text.

I unfortunately cannot find scans of the original 1904 or 1912 print publications anywhere, although they should be public domain.

[Alan Bunning]

Welcome to my world. I found the same thing was true when examining copies of Westcott and Hort, Stephanus 1550, etc. In fact, NOT ONE of the electronic texts I found on the Internet and in Bible programs turned out to be correct for any text, including examples where the text was taken directly from the author’s website. I am interested in this, because I was thinking about adding the 1904/1912 text to my website, and would have to do the exact same thing. I too have collected the different electronic versions and noticed that there were differences, but just have not got around to trying to iron them all out yet for this text. I did find a pdf scan of the 1904 text, but haven’t found one for the 1912 text yet. Perhaps we can talk off-line, because if you can sort it all out, you would save me all of the trouble.

[Jonathan Robie]

I maintain a version of the Nestle1904 here, and would welcome collaborators:

https://github.com/biblicalhumanities/Nestle1904/

It includes morphological analysis, and we also have syntax trees and other datasets for it here:

https://github.com/Clear-Bible/macula-greek

[Alan Bunning]

Just to be clear, we were talking about the differences in the 1904/1912 Patriarchal text, which is different than the 1904 Nestle text, which is a different animal.

[Jonathan Robie]

Just to be clear, we were talking about the differences in the 1904/1912 Patriarchal text, which is different than the 1904 Nestle text, which is a different animal.

Yes. Both in 1904. Sorry to be confusing.

I also have a copy of the 1904 Patriarchal Text that is probably close to the eBible text. I would love to see collaboration around that text too. One complication: the Hellenistic Bible Society really does think they still have copyright to that text, even though the copyright has clearly expired.

[doday]

Thanks, all! Yes, I will DM you, Alan. I'm working with a team that is releasing an updated (public domain) version of the Robinson-Pierpont GNT with comparisons to other printed editions. We have written a bit of code to do collations with associated Unicode decomposition, removal of punctuation (but retention of brackets or other indicators of uncertainty), fixing code points (e.g., GOA uses the ohm symbol instead of omega in many places, and certain ASCII characters at times instead of the Greek code point), etc. to make collation more automated. We have a visual editor where we can tag certain variants to create footnotes in our edition (it outputs to LaTeX), but are storing our tagging in a SQLite database we intend to share on GitHub. we are ignoring things like movable nu, elision, minor orthographical variations, etc.

But we are running into problems with trusting the source texts we are comparing to, which defeats the point a little since our goal was to compare to such printed editions instead of manuscripts. Also, Alan, I only became aware of your apparatus and collation tools a couple weeks ago when posted here—awesome tools!

[doday]

For visibility, I ended up posting a little longer version of the above (with a couple more examples and some additional editions compared) on my blog (with a link back here) in case someone stumbles across it in the future: Would the real 1904/1912 Patriarchal Greek New Testament (PATr) please stand up?