From the Library: GA 2907


By: Andrew K. Bobo

The Center for the Study of New Testament Manuscripts (CSNTM) Digital Library contains hundreds of Greek NT manuscripts, each with its own story to tell. In our “From the Library” series, we will feature individual manuscripts from our collection in order to showcase their unique beauty and importance. This is part of CSNTM’s mission to make NT manuscripts accessible for everyone. 

The manuscript now known to New Testament scholars as Gregory-Aland 2907 was ‘discovered’ by CSNTM almost a decade ago. The Center became aware of this manuscript through a route full of intrigue. A friend of the Center—whom we have never met—was on the lookout in his country for uncatalogued New Testament manuscripts. After some super-sleuthing, he was able to locate the owner, a private collector, and he put them in contact with the Center. CSNTM then partnered with the owner to digitize the manuscript and make the images available online. This was highly significant, since GA 2907 is a first-millennium witness to the text of the Gospels, and its witness is only now being taken into account by scholars.

From the work of Darrell Post, who did a collation of the entire codex, we have learned that the text of GA 2907 very closely resembles the Majority Text. According to Post, the original scribe was careful and “committed very few unforced errors in the copying of this manuscript.” The original scribe wrote “very neatly” and was even neat in correcting the text, leaving little or no trace of the mistakes in places where the text has been scrubbed and rewritten.

In contrast to the scrupulous work done by the original scribe, the extensive repairs and ‘corrections’ of a later scribe appear clumsy and at times even bizarre. As Post notes, the later corrector’s attempts to retrace over the original scribe’s writing often did “more harm than good.” The corrector “sometimes left alone faded brown letters, and at other times traced over perfectly legible letters.” This is reminiscent of what a later copyist did to the text of Codex Vaticanus, although 2907’s ‘corrector’ was not in the same league as Vaticanus’s corrector.

A good example of such corrections comes towards the very beginning of the manuscript in Matthew 1.

Matthew 1 in GA2907

You can see the original scribe’s writing in the top half of the page. Then the later scribe’s retracing begins in dark black ink on the bottom half. The retracing skips some letters and does not trace well over others, ignoring the form of the letters in some cases.

Here is another example from a few leaves later at the end of Matthew 2 and beginning of chapter 3.


In the fourth line from the bottom, you can even see an instance where the corrector’s re-written line completely departs from the original scribe’s. The reasons for the corrector’s sloppy work are unknown, but they illustrate the fascinating and complicated histories that manuscripts can have. GA 2907 was obviously well worn from centuries of use, with someone even going through the trouble of trying to make the manuscript usable again after the original work had become damaged and faded.

 Gospel Titles in GA 2907

An idiosyncratic feature of GA 2907 is the title given to each Gospel. In the three extant titles by the original hand, shown above, the scribe includes an extra preposition within the traditional formulation. Typically, the title is written: “The Gospel According to Mark.” But in 2907, the scribe wrote “The Gospel From the According to Mark,” adding the Greek preposition ek, or “from,” to the title.  This way of writing the title would have been typical of a lectionary, where the manuscript contains selections from a Gospel rather than the entire text of a Gospel. It seems likely that this scribe inadvertently used lectionary titles here out of habit, perhaps because the scribe usually copied lectionaries rather than minuscules.

 Missing Pericope Adulterae in the Gospel of John

Another idiosyncrasy of GA 2907 is how the manuscript deals with the pericope adulterae, the story of the woman caught in adultery (John 7.53–8.11). The text of GA 2907 follows a well-known group of manuscripts, referred to as “family 13.” This family of closely related manuscripts places the pericope adulterae in an odd place—towards the end of Luke rather than in John’s Gospel. GA 2907, however, appears to differ from its close relatives in some interesting ways. Instead of inserting the material from John after Luke 21.38, GA 2907 inserts the material just after Luke 23.33. The material inserted is from John 7–8, but curiously the pericope adulterae itself is not included. Instead, the manuscript’s original hand moves directly from John 7.52 to 8.12 without any break. There is writing in red ink just between these two verses that could indicate that something is missing, but it is unclear. So although the manuscript has a type of text which we would expect to contain the story, instead it is missing entirely.

 Missing the Pericope Adulterae in John

A close-up of the transition between John 7.52 and 8.12.

GA 2907 illustrates how CSNTM is contributing to scholarly work on the Greek NT. During the last 15 years, we have discovered scores of manuscripts which were previously unknown and uncatalogued. This came about through our collaborations with manuscript owners to make their collections available freely on our website, which then allows the manuscripts to be consulted by the editors of critical editions of the Greek NT. GA 2907, now less than a decade after CSNTM discovered it, was cited as one of the manuscripts consulted in a recently published critical edition, The Greek New Testament, Produced at Tyndale House, Cambridge(THGNT). THGNTand critical editions like it are the base texts used for Bible translators, whose work will soon be in the hands of readers worldwide.

It is important to remember that even today, there are still manuscripts that lie undiscovered and their treasures unexplored. We want to find them. We hope that you will partner with us to discover the undiscovered in order to make it available for all.