A Less Studied Isaiah Scroll

In my own personal study of Isaiah, I’ve come across a less known and less studied Dead Sea Scroll, 4QIsao (4Q68), that preserves some insightful, even peculiar details. It dates paleographically to 100-50 BC which makes it a little younger than the more popular 1QIsaa.

In this photo of 4QIsao, you’ll notice several yellow underlines, two red underlines, and one blue underline. Seven yellow underlines draw our attention to some peculiar uses of a final mem used in non-final position. From my experience, this feature is seldom evidenced in the Dead Sea Scrolls. Has anyone seen a comparable text?

The red underlines refer to two orthographic variants. In these two instances, the scribe of 4QIsao transcribes the more common form of “night” לילה for the MT’s less common form ליל. The difference here regards spelling and that of only one letter – a heh – and probably represents a facilitating approach to copying the text since the form of 4QIsao is the more common spelling.

Finally, the blue underline draws our attention to an ancient textual variant. Here is a breakdown of the textual evidence in MT, 1QIsaa, and 4QIsao.

MT (Isa 14:31)		1QIsaa (Isa 14:31)		4QIsao (Isa 14:31)
וְאֵין בּוֹדֵד בְּמוֹעָדָיו	And there is not one who separates in his ranks	ואין מודד במודעיו	And there is not one who measures (out punishment) for his kinsmen	ואין] בודד בםידעיו	And there is not] one who separates among his friends

The witnesses are all genetically related and concern three interchanges: 1) interchange of bet (MT/4QIsao) and mem (1QIsaa), 2) interchange of vav (MT/1QIsaa) and yod (4QIsao), 3) and transposition of dalet and ayin (MT vs. 1QIsaa and 4QIsao). Each of these interchanges is common in the Jewish Square Script. The MT is probably the more original reading with the two Dead Sea Scrolls trying to make sense of the MT’s difficulty. The difficulty of the MT is further evidenced in the Septuagint’s confused translation of καὶ οὐκ ἔστιν τοῦ εἶναι. Since at least the second century BC, the difficulty of the MT appears to have puzzled scribes and translators. 

Overall, this manuscript stands out for its peculiar usage of final mem in non-final position, a feature that marks it as exceptional among Dead Sea Scrolls. The rest of the scroll is typical for Qumran texts, fragmentarily preserving a biblical text that aligns very closely with Codex Leningrad, matching it at 97.77% when excluding two spelling differences. While it contains one insightful textual variant at Isaiah 14:31, this variant appears to interpret the MT in light of its context. In a small but meaningful way, 4QIsao affirms the fidelity of the Masoretic tradition. Its text is nearly identical to it, and when it diverges, it does so in predictable ways.  

GA 2136, Another Copy of a Printed Edition

As I mentioned in my last post, I’m in the process of working through all the Pauline manuscripts in the Liste, in order to get a precise count of which manuscripts contain some portion of Romans. Of the 651 extant, or possibly extant, manuscripts currently classified as minuscules that contain some portion of Romans, I’ve so far been able to examine 632.

However, based on a lead that Katie Leggett sent in response to a forum query, we can now remove one more of those unexamined witnesses as a copy of a printed edition. 

GA 2136 (Moscow, Syn. gr. 472, Pinakes Diktyon 44097), is a Greek-Slavic bilingual from the 17th century that contains the entire New Testament. The VMR has microfilm images folios 96-129 (Mark 1:1-Luke 1:10) and folios 185-226 (John 1:1-21:25). However, the Moscow State Historical Museum has also made available a few color images, including the beginning of Revelation. (Thanks to Katie for pointing these out to me). 

These images, together with the microfilms already available in the VMR, make it clear that the Greek text in this ms has been copied from a printed edition and, as such, should be removed from the Liste. Four features are particularly significant. I've included screen shots

1. As usual, the biggest giveaway is that it doesn't include nomina sacra. Unless there are very strong reasons to counteract it, this alone would be enough to demonstrate that a manuscript has been copied from a printed edition. But this is not all. 
   
   
   
2. Another giveaway is that the manuscript uses indentation instead of ekthesis. 
   

   
   
3.  Perhaps the most striking feature, however, is that the manuscript includes modern verse numbers. These verse numbers could not have been added later from a secondary source because the indentation follows them precisely.  
   
   
4. A final feature, not, of course, decisive on its own is that (as shown in the image linked to above) Revelation is immediately preceded by Jude, which is a rather unusual in Greek manuscripts, but standard in early printed editions. 

While I have illustrated these points from the color images made available by Moscow, the three primary points made above are all also visible in the microfilm images of the Gospels available in the VMR. 

In other words, it is not simply that Revelation has been added from a printed edition. The Greek text as a whole has been copied from it, and as such should be removed from the Liste. On another note, it is interesting that this manuscript contains what appears to be text-critical sigla like brackets (see below from the end of Jude)—it would be interesting in its own right to work out what these refer to. 

Darrell Post on a New Lectionary Group

The following is a guest post by Darrell Post who has been researching John 11 for many years (see here and here and here). In this connection I also want to mention that Post has done a great job with entries on Greek New Testament manuscripts on Wikipedia.

  

A Newly Found Lectionary Group?
by Darrell Post

New Testament textual critics have lamented the status of lectionary studies for generations. Despite the large quantity of these liturgical works, there are more continuous-text manuscripts and many of these remain to be fully studied. Lectionaries have been perceived as secondary, not existing in the early centuries of the church, and therefore further removed from the initial text. Most lectionaries are presumed to align closely with the Byzantine text and would therefore add very little to what is already known.

 

Happily, research into lectionaries has been increasing in recent decades, so when I embarked on a project five years ago to collate manuscripts that include John 11, I wanted to involve lectionaries in my work.  Compared to continuous-text manuscripts, fewer lectionaries have been digitized and made available on the internet. Far fewer have had their images indexed with the Scripture references. Locating a specific New Testament portion can be a challenge given the varying lectionary formats. I was fortunate with John 11 as the lesson that includes verses 47-54 appears early in most lectionaries and follows the lengthy lesson about the man born blind in John 9. The other lesson, John 11:1-45 is normally found just a few lessons past the artwork signaling the start of the Gospel of Mark lessons. But the easy way to find it was by scrolling through the images looking for the rare use of the decorative pi that stands out for starting the lesson of John 12, as this lesson is separated from John 11 by only one lesson from Matthew.

 

A year ago, my work was finished with continuous-text manuscripts, having collated 1,726 of them against John 11. One of them, GA-345, caught my attention in June 2021. Upon examination, the text presented a strong resemblance to the pattern of variations found in the papyri and early majuscules. Wisse had found the text of Luke in GA-345 to be within the Byzantine tradition[1], and checking the Manuscript Clusters tool on the INTF’s website[2], little returned in the search of GA-345 against the samples taken from John. I presumed the text of John 11 within GA-345 to be an anomaly. 

 

As my work turned to the Lectionaries, my initial survey found a possible 881 that might include John 11 and have images available to study. I have now completed the work on 382 of the 881. Lectionary manuscripts might naturally be presumed to have been always created from other lectionaries, meaning they would have developed their own patterns of variation, distinct from continuous-text manuscripts. However, Burns found in 1982 that L574 belonged to Family 13.[3] 

 

I have found comparable results so far in John 11. Several lectionaries include readings distinct to the Π family, including L2023, a match to the text of Π members GA-268 and GA-787. After the launch of the Kr group among continuous-text manuscripts, at least a dozen lectionary texts were copied in conformity to Kr readings.
 

Just recently I have collated John 11 in four lectionaries that surprisingly were found to share the variation pattern of continuous-text GA-345. These five manuscripts present an early form of the text of John 11:

345, 11th, Milan, Biblioteca Ambrosiana, F. 17 sup.
L466, 12th, Grottaferrata, Biblioteca Statale del Monumento Nazionale, A. α. 11.
L518, 13th, Messina, Biblioteca Regionale Universitaria, S. Salv. 95.
L534, 12th, Vatican, Biblioteca Apostolica Vaticana, Barb. gr. 448.
L546, 10th, Vatican, Biblioteca Apostolica Vaticana, Vat. gr. 781.

I have not found anything written about these lectionaries and the readings they share in John 11, although I am still looking and invite anyone to point out something I have missed. Jordan did not select any of these manuscripts for analysis in his dissertation on the lectionary text of the Gospel of John.[4] 

 

Each of these manuscripts needs to be fully collated to see the extent of these distinctive readings. L466 has lost the opening pages including the lesson covering 11:47-54. L518 and L534 follow the Majority Text in 11:47-54, but match GA-345 in 11:1-45. L546 follows GA-345 in both lessons, indicating the potential for more parallels.  

 

Morrill found that GA-345 has 96.1% agreement with the Majority Text with no close relatives in John 18, so this manuscript was not selected in his list of recommended manuscripts to include in the ECM edition of the Gospel of John.[5]

 

Given the text presented by GA-345 and these four lectionaries, at least in John 11, future critical editions should consider including the testimony of these witnesses. Most importantly, this find demonstrates that lectionaries are worthy of study with the potential for more surprises to be found that are right now hiding in plain sight. 

 

The chart below shows a comparison from John 11 of 345, L466, L518, L534, and L546 against well-known manuscripts:

---

[1] Frederik Wisse, The Profile Method for Classifying and Evaluating Manuscript Evidence (Grand Rapids: Eerdmans, 1982), 110.

[2] http://intf.uni-muenster.de/TT_PP/Cluster4.php

[3] Yvonne Burns, “A Newly Discovered Family 13 Manuscript and the Ferrar Lection System.”

In Studia Patristica, vol. XVII, part 1, edited by Elizabeth A. Livingstone, 278– 289.

Oxford: Pergamon Press, 1982.

[4] C.R.D. Jordan, ‘The  Textual Tradition of the Gospel of John in Greek Gospel Lectionaries from the Middle Byzantine Period (8th-11th Century)’ (PhD; Birmingham, 2009), 48-49.

[5] Michael Bruce Morrill, ‘A Complete Collation and Analysis of all Greek Manuscripts of John 18’ (PhD; Birmingham, 2012), 169, 386.

Correcting a Dead Sea Scroll

Correcting a digital text today is simple and neat. I simply highlight the text to be deleted and press my “delete” button on my keyboard. When I notice a mistake right away, I can simply press the undo button to erase my last input. The phenomenon of simple and neat corrections is a modern reality, not an ancient one. In this post, I’d like to highlight some techniques scribes used to correct the biblical text. 

Deleting Text
When ancient scribes deleted text, they could use “cancellation” dots. Both examples below come from 1QIsaa. The second picture is quite peculiar since the initial scribe corrected the text, making the verbal stem explicitly qal, only for the same scribe or a future scribe to delete the correction with dots. The result of the cancelation dots is the verbal stem once again becomes niphal.

Scribes could cross out the material to be deleted. Here is an example from 1QIsaa.

Another option was to bracket the content in parentheses. Here is a partially extant example from 11QpaleoLev. Only the final bracket is preserved.

Scribes could also scrape the ink off the writing surface (i.e., erasure). The scribe of 1QS often resorts to this technique.

Adding Text
By far, the most common way to add originally omitted text was with a supralinear correction (i.e., adding the material above the line). This type of intervention is common, especially in 1QIsaa

Substituting Text
Scribes often needed to correct the text by substitution. They could do this by combining the features above such as cancellation dots plus a supralinear correction. Reshaping an errant letter was another option. This technique often resulted in recognizable but peculiar letters. Here are two examples from 1QIsaa.

This second example is especially important since the reshaped letter is probably an aleph. Most likely the scribe was in the process of written ארצ (with a non-final sade) based on the context but realized his mistake and corrected the text to read השמים.

Correcting ancient manuscripts could get messy as these above examples prove. It is important that textual critics and those concerned with the transmission of Bible remind ourselves of this reality. What is easy and neat for us, was not always neat for ancient scribes copying the Bible. 

The pictures come from https://www.deadseascrolls.org.il/?locale=en_US and http://dss.collections.imj.org.il/.

New York Times Review of Westcott-Hort

Exactly 143 years ago to the day, America's paper of record (aka The New York Times) published a review of a recently published introduction to a Greek New Testament by two Cambridge scholars named Westcott and Hort. It is not signed and my online source (ProQuest) does not give an author, but your best guesses are welcome in the comments.

Appreciating 11Q5: A Comparative Look at the Great Psalms Scroll

11Q5 (11QPsa), known as the Great Psalms Scroll, is on display at the Ronald Reagan Library in Simi Valley, CA. This exhibition has been made available by the Israelite Antiquities Authority and the sponsorship of The Lowell Milken Family Foundation. You can learn more about the exhibition here. I had the opportunity to visit the exhibit on Friday with my family and found it to be an enriching experience.

In this post, I’d like to discuss three (among many!) important features of this manuscript by way of comparing it to the more popular 1QIsaa. For a discussion on the textual nature of the manuscript, see my prior post here.

1) Writing Precision and Formatting
One of the most noticeable differences between 1QIsaa and 11QPsa is the approach to the writing block. A writing block is the rectangular shape created by the top and bottom horizontal lines and the right and left vertical lines (See Tov’s Scribal Practices pp. 82-108). The scribe of 11QPsa writes within the writing block while the scribe of 1QIsaa often transgresses it. Notice that the scribe of 11QPsa only goes beyond the left vertical line twice in col 7.

Compare this column to column 2 of 1QIsaa. Here the scribe transgresses the left vertical line in almost every line.

The lack of concern for the writing block is perhaps most pronounced where the scribe begins a word to the left of the vertical line. Although only one letter of the word remains (due to deterioration of the manuscript), the letter is clearly to the left of the writing block and the beginning of a new word. See an example of that here.

Similarly, sometimes the scribe of 1QIsaa begins a word and realizes he cannot complete it in the remaining space. In these instances, the scribe has recourse to some “less than ideal” procedures. For example, here he begins a word, realizes he cannot finish it, stops, and transcribes the word in its entirety at the start of the next line.



In another situation, the scribe crams the final portion of the word (a pronominal suffix) above the line.



This feature of writing in the writing block reflects to some degree the skill of the scribe. The scribe of 11QPsa was more skilled than the scribe of 1QIsaa, at least regarding the spacing of words.

2) Degree of Scribal Intervention
Both manuscripts display a different level of scribal intervention. 1QIsaa has an instance of scribal intervention every 4 lines while 11QPsa has an intervention every 9 lines. (See again Tov's Scribal Practices, pp. 332-335). Neither statical is very impressive to be sure. 

A brief survey of the columns indicates that the scribe(s) of 1QIsaa intervened in the text much more frequently than the scribe of 11QPsa. Here is a column from each respective manuscript for review.

 

3) The Paleography
The paleography is also different. Although not uncommon among biblical manuscripts, it is exciting to see the divine name written in paleo-Hebrew in 11QPsa rather than the Jewish-Aramaic Square Script.

Next, the script is clearly from different eras. Some features of 1QIsaa include inconsistent use of final letter forms in final position (and even final letter forms in non-final position). The letter forms of 1QIsaa are not consistent in size. The script of 11QPsa, however, uses final/non-final letters and represents the letter size consistency. Here are some specific letters for comparison.

See the different form of the samek. In 1QIsaa, the form is open at the bottom.



See the different size of the ayin. In 1QIsaa, the form is more truncated.

Notice the backyard tick at the top of the zayin in 11Q5. It is more pronounced than in 1QIsaa (usually).

Notice the length of the downward stroke of the yod and the nature of the letter's head. The letter form is longer and more angular in 11QPsa.

There are several fascinating features of 11QPsa. These are just three. I hope this post stimulates more work on this important manuscript and readies visitors for their visit to the Reagan Library Your visit will not disappoint.