The Date of P75 and the Length of Kollemata in a Papyrus Roll

Image of Subatianus Aquila order to release a prisoner mentioned by Turner, "The Terms Recto and Verso," page 61 (see original image here)

In a newly published article in Novum Testamentum, Brent Nongbri argues for a secure fourth century date for P.Bodmer XIV-XV (P75) based on the length of kollemata used in the manufacture of the roll the individual codex pages were cut from. The article is open access and available at the following link.

Brent Nongbri, The Construction and Date of P.Bodmer 14–15 (P75): The Length of kollemata as a Criterion for Dating Papyrus Codices", Novum Testamentum 68.4 (2026): 518-538.

Here is the abstract of the article.

This article describes an underutilized criterion for assigning dates to papyrus codices, the presence of very long sheets (kollemata) in the rolls from which the bifolia of papyrus codices were cut. In the 1970s, Eric Turner argued that a shift in the production of papyrus rolls took place in the fourth century, at which point rolls began to be manufactured using kollemata much longer than the 10–30 cm kollemata normally used in the early Roman imperial period. The present article briefly discusses the manufacture of papyrus rolls, gathers evidence for longer kollemata, and places this evidence in conversation with P.Bodmer 14–15 (P75). The date of this manuscript has been the subject of some debate in recent years, and consideration of the length of its kollemata adds another, perhaps decisive, piece of evidence that this codex was produced no earlier than the fourth century.

Overall, the article is an excellent piece of detective work that summarizes the ancient manufacturing process of papyrus rolls mentioned by Pliny the Elder in his Natural History, 13.74-82. Nongbri assembles the very limited evidence that is preserved from antiquity. He examines several codices that were constructed with pages cut from rolls that were manufactured by pasting very long sheets of papyrus together.

Papyrus Bookroll made up of individual papyrus pages glued together, each seam, or kollesis is highlighted in red (illustration adapted from Turner, "The Terms Recto and Verso," page 14)

Papyrus bookroll illustrating how each codex bifolium were cut without regard for the kolleseis (illustration adapted from J. Scherer, Extraits des livres I et II du Contre Celse d’Orige`ne d’apre`s le Papyrus no. 88747 du Muse´e du Caire, Cairo, 1956)

Nongbri provides several examples of codices made from papyrus bookrolls with wide kollemata that can be securely dated to the fourth century and after. From these examples, and going off some statements by Turner, Nongbri makes a broader argument, that codices manufactured from rolls with wide kollemata can be dated to the fourth century or after based on this feature alone. Turner's comments about the width of kollemata are found in the following work.

Eric G. Turner, "The Terms Recto and Verso. The Anatomy of the Papyrus Roll" in Actes du XVe Congrès International de Papyrologie (Brussels: Fondation Égyptologique Reine Élisabeth, 1978), page 62.

At first glance, these observations by Nongbri seem to be solid objective evidence that P.Bodmer XIV-XV (P75) must be dated into the fourth century. However, there are a few problems with this argument that nulify these claims.

First, the date ranges of several of Nongbri's examples actually span into the second and third centuries. 

PGM I. Though it has been palaeographically dated from the third into the fourth or fifth centuries, recently, in a detailed examination using securely dated examples, Alberto Nodar has assigned this fragment to the third century (see the full open access article here).

PGM II + PGM VI. This magical papyrus has been palaeographically dated from the second to the third century (see summary here).

The Crosby-Schøyen Codex. This Coptic codex has been dated using Carbon 14 with a date span ranging from the third to the fourth century (see article here). It has also been palaeographically dated from the second to the sixth centuries.

Second is the passage in Pliny the Elder's Natural History 13.74-82 concerning the manufacturing of papyrus. I won't quote the passage in full, however, one section in 13.80 states, 

Claudius also increased the width of the sheet, making it a foot across. There were also sheets a cubit (44 cm) across called 'macrocola,' but examination detected a defect in them, as tearing off a single strip damaged several pages. On this account Claudius paper has come to be preferred to all other kinds, although the Augustus kind still holds the field for correspondence; but Livia paper, having no quality of a first-class kind, but being entirely second class, has retained its position.

erat et cubitalis macrocollis, sed ratio deprehendit vitium, unius schidae revulsione plures infestante paginas. ob haec praelata omnibus Claudia, Augustae in epistulis auctoritas relicta; Liviana suam tenuit, cui nihil e prima erat, sed omnia e secunda.

There are a few issues with this quotation from Pliny the Elder. Adam Bülow-Jacobsen, "Writing Materials in the Ancient World," in The Oxford handbook of Papyrology (pages 5-6) notes that there are many problems with Pliny's statements that betray his lack of firsthand knowledge of the process. One in particular concerns Pliny's comments that the wider rolls with a kollema width of one cubit length were not favored for bookrolls. This larger roll he referred to as the 'macrocola' (Latin: macrocollis) was not preferred because a strip of papyrus might come loose from the page and potentially ruin a large section of writing. Nongbri seems to make much of this fact and suggests that this may be the reason these rolls with extremely wide kollemata of a meter or more were made specifically for codex leaves. The problem is that despite what Pliny says here, the macrocola rolls were used for and seemed to have been favored for extralarge deluxe volumes. Cicero preferred these larger rolls when he wanted a particularly fine copy of a book made. To Atticus he wrote,

"I am sending you the same composition more carefully revised, indeed the original copy, with plenty of additions between the lines and corrections. Have it copied on large paper (macrocollum) and read it privately to your guests; but, if you love me, do it when they are in a good temper and have had a good dinner, for I don't want them to vent on me the anger they feel towards you." (Att. XVI.3)

In another place, Cicero wrote to Atticus,

"But what on earth is the reason why you are so frightened at my bidding you send the books to Varro on your own responsibility? Even now, if you have any doubts, let me know. Nothing could be more finished than they are. I want Varro, especially as he desires it: but, as you know, he is "a fearsome man; the blameless he would blame." I often picture him to myself complaining of this perhaps, that my side in the books is more fully defended than his own, though I assure you, if ever you come to Epirus, I will convince you it is not. For at present I have to give way to Alexio's letters. However, I don't despair of winning Varro's approval; and, as I have gone to the expense of a large paper copy (macrocolla), I should like to stick to my plan." (Att. XIII.25)

One could also mention Catullus in his Poem 22 where he makes reference to the use of "royal paper." This is often thought to be the same type of large format 'macrocola' roll, mentioned by Cicero and Pliny (see McDonnell, "Writing, Copying," page 479, especially footnote 44). This fact, that these larger rolls were actually used and preferred for books, brings some doubt on the rest of what Pliny states. It may be that other, larger rolls were being produced and used as well, especially in non-literary contexts (more on that below).

The third problem with Nongbri's argument has to do with Turner's own comments surrounding these codices cut from rolls with wider kollemata. Turner does state that a shift happened in the fourth century, but actually suggests that the phenomenon began in the third century.  Here are Turner's observations.

"6.2 However, starting in the fourth century a change in the process of manufacture must have taken place, for one begins to find sheets of greater breadth. . . . From inspection of the make-up of the Coptic codices from Nag-Hammadi J. M. Robinson has concluded that kollemata of "more than a metre" in breadth were used to make the rolls then cut up for the codices found there (Robinson 1975 p. 19; Robinson 1977). The quality of these sheets is far from fine: indeed they could only have been made by utilizing a length of stalk of the papyrus plant that manufacturers of an earlier age would have discarded (a consideration which incidentally reinforces the interpretation of "principatus medio" in Pliny n.h. xiii 74 offered by Bülow-Jacobsen 1976). Nevertheless it must be accepted as established that such a change in manufacturing standards and methods took place (possibly as early as the third century A.D.)." (Turner, "The Terms Recto and Verso," page 62).

To Nongbri's credit, he does bring this fact to light in his article, however, he states confusion over why Turner would suggest a third century date for the transition because "[t]he papyri he [Turner] mentions in the immediate context all date to the fourth century or later: (page 532, note 46)." What is striking, is that on the previous page, Turner discusses three examples of papyri with very large kollemata that date in the second or third centuries; these are given below (Turner, "The Terms Recto and Verso," page 61).

P. Oxy 37.2806 (see website here)

P.Oxy 37.2806. This is a Greek comedy that has been palaeographically dated to the second or third century (see website here). Only a fragment of the roll is preserved but in the 32.5 cm wide segment no kollesis was discovered and it is possible that, depending on how one places other preserved segments, the minimum width of the kolema was 40 cm. This is either a "macrocola" roll with a kollema of 44 cm as mentioned by Pliny, or this could have been one of the larger rolls (like those examined by Nongbri) used for a literary bookroll, we simply do not know.

P. Strasbourg 1.6. This is a documentary roll with 36.5 cm measured between the kolleseis and is securely dated to 244-261 CE (see website here).

P.Berol. 11532. This is a famous decleration that the Prefect of Roman Egypt, Subatianus Aquila wrote to Theon, the Governor of the Arsinoite nome, ordering the release of a man named Niger, son of Papirius (see the website here, and also here). The width of this sheet is 32 cm with no kollesis, so it had to have been cut from a roll made from pages at least this width.

These examples that Turner gives from the second and third centuries are probably what gave him pause and pushed him to suggest a shift in the third rather than the fourth century. When these examples from Turner are added to those second and third century examples given by Nongbri above, Turner's third century date is reinforced.

Finally, Turner originally suggested these rolls with wider kollemata may have been manufactured specifically for the construction of codices and Nongbri argues this in the article. However, perhaps these extremely long rolls were originally manufactured for use in a slightly different context. If we take a hint from the two examples given by Turner, (P. Strasbourg 1.6 and P.Berl. 11532) it may be that rolls made from extremely wide sheets were originally manufactured to be used for documentary purposes. The wide kollema would allow for varying sizes of individual sheets to be cut (such as the very wide prisoner release by Aquila) without having to take into account a kollesis. Along these same lines, many scholars have noted that the codex format had its origins within a "documentary" rather than a "literary" context (see the comments of Bülow-Jacobsen, "Writing Materials in the Ancient World," page 23). It may be that the rolls with wider kollemata were already being made for documentary use, whether to cut individual pages to be used as letters, lists, contracts, etc., or for codices used in a documentary context (See Galen, "On the Avoidance pf Grief," 31-33). It seems natural that some codices used for "literary" works may have still been cut from these rolls intended for "documentary" purposes. Note Turner's comments on these extremely wide rolls that the Nag-Hammadi codices were cut from, that the "quality of these sheets is far from fine" ("The Terms Recto and Verso," page 62). Perhaps it is because these codices were cut from very wide, poor quality papyrus intended for documentary needs. According to the very scant evidence discussed here, these rolls with wide kollemata began to be used for codices some time in the third century, but likely existed in documentary contexts much earlier. As the codex format began to be used in greater frequency in the fourth century, it may be that the market for these rolls with wider kollemata began to increase. Therefore, the fact that P.Bodmer XIV-XV (P75) was manufactured from a roll with extremely wide kollemata, according to the evidence, fits well within a third century context.

 _______________________

Myles McDonnell, “Writing, Copying, and Autograph Manuscripts in Ancient Rome.” The Classical Quarterly, 46.2 (1996): 469 - 491.

 

Forthcoming in The Journal of Theological Studies: Moses as a Symbol of the Crucified Jesus: An Early Reference to the Staurogram?

Early on in my studies, I came across Larry Hurtado's book The Earliest Christian Artifacts: Manuscripts and Christian Origins. It was deeply informative and fascinating to learn of all the features that were characteristic of Christian books; the codex, readers aids, and the nomina sacra (among other distinctives). I used his work as the basis for a detailed adult Sunday School class that I gave at my home Church over a period of several weeks that covered the information found in The Earliest Christian Artifacts. I had the opportunity to share this with Hurtado one time at SBL and he was thrilled. He said that one of his goals for writing that book was to inform the every day Christian of the rich heritage of the manuscript tradition of the New Testament. 

Hurtado was also well known for his work on the staurogram, to which he dedicated a chapter in his book. I too became fascinated with the nomina sacra (thanks to Hurtado), and in particular this curious tau-rho ligature. I was convinced by Hurtado's arguments that this device was used as a sort of crude pictogram of the crucified Jesus. Eventually this curiosity led to me writing an article on the origin of this device. After an earlier draft version was rejected by JBL, I submitted my manuscript to The Journal of Theological Studies and I recently learned that it was accepted for publication. I wish that Hurtado was still with us because I would love to talk about the staurogram with him over a cup of coffee.

I am not sure when the article will appear in print. Until that time, here is the title and an abstract to whet your appetite.

Moses as a Symbol of the Crucified Jesus: An Early Reference to the Staurogram?

Abstract:
This article investigates the origins of the Christian use of the staurogram, a ligature formed by the superimposed Greek letters tau and rho, as a symbol of the crucified Jesus, in three early New Testament gospel codices; P66, P75, and P45. Evidence is presented from the Epistle of Barnabas 12.1-3, which interprets Moses’s outstretched arms during the battle against Amalek as a prefiguring of Jesus’s crucifixion, demonstrating that the veneration of cruciform imagery was already present within the imagination of some Christian communities. This suggests that all the necessary ingredients and catalysts were present for the appropriation of the tau-rho ligature at the beginning of the second century. The presence of the staurogram in the Beratius Nikatoras inscription, several Christian gemstone rings, Codex Sinaiticus, and the “Jung Codex,” and the nomen sacrum evidence from a recently discovered Frankfurt Silver amulet, are used to construct a chronological framework for the adoption of the staurogram in the New Testament gospel papyri. This article argues that there is good reason to believe staurograms could have appeared in manuscripts of the canonical gospels as early as the end of the second century and were adopted as standalone pictograms by the fourth century

Key words: staurogram, nomina sacra, Epistle of Barnabas, Ephrem the Syrian, Frankfurt Silver Amulet, P45, P66, P75, Codex Sinaiticus

 

P.Bodmer II (P66), a Marginal Correction, and the Useful Life of Books

While re-reading James Royse's discussion of P66, I came across his brief mention of a correction at John 13:19a. It seems that the original copyist skipped over an entire line in their exemplar and omitted the following phrase in red, 

ἀπ᾽ ἄρτι λέγω ὑμῖν πρὸ τοῦ γενέσθαι, ἵνα πιστεύσηται ὅταν γένηται ὅτι ἐγώ εἰμι.

The interesting feature about this correction is that it appears to be the only correction not made by the original copyist of the manuscript. Here are Gordon Fee's comments in full.

"However, there is one correction which seems clearly to be the work of a second hand: the addition of απ αρτι λεγω υμιν προ at 13:19. The square μ and υ simply demand a second hand: in the original scribe’s hand (even in the corrections which are obviously his) these letters are always well-rounded. Although this particular hand does not seem to be clearly in evidence at any other point, it does indicate that a second hand has had access to the MS." (Fee, Papyrus Bodmer II, 59)

The script used in P66 has been classified by Turner as "informal round" which has letters that "are rounded but also seem flattened" (Turner and Parsons, Greek Manuscripts of the Ancient World, 21). Orsini and Clarysse have placed the script of P66 into their "Alexandrian Stylistic Class" which is characterized by letters that are "unimodular and looped, and the strokes end in apices (in the lower parts) and small hooks (in the upper parts); sometimes curves and flourishes are added at the end of letters." (Orsini and Clarysse, "Early New Testament Manuscripts and Their Dates," 458).

 

P.Bodmer II showing correction at John 13:19

When examining the interlinear correction at John 13:19 in closer detail, the difference in the two scripts is apparent. As Fee noted in the quote above, the μ and υ are markedly different, the correction being more angular and less looped and curved than the letters of the main text. Also note the difference in shape and the flow of the pen strokes in the way the alpha, mu, upsilon, and omicron are formed (these letters are highlighted in red below).

Closeup detail of John 13:19 correction in P.Bodmer II (P66). Compare the highlighted letters in the original script with those of the interlinear correction at the top.

Also, added to the difference in letter shapes is the difference in ink. It is difficult to see the details in the gray scale image, but it appears that the ink of the marginal correction is darker and the ductus is sharper with less fading from the ink. Again, it is difficult to tell from this image, but it appears to be a different ink that was used in the correction as compared to the main body of text. When comparing the script of the correction at John 13:19 with the more formal rounded biblical majuscule of the fourth and fifth centuries, clear parallels can be seen.

Closeup detail of John 13.19 in Codex Sinaiticus (top), Codex Alexandrinus (Second from top), Codex Vaticanus (Second from Bottom), and the Correction at John 13.19 in P66 (bottom).

I aligned the same text from John 13:19 with the text from three other codices in order to illustrate the close similarities between the style of script used in these manuscripts (in Sinaiticus and Vaticanus, the text spread across two lines so I altered the image so that the text from these manuscripts would appear horizontally together for comparison purposes). When looking at the overall letter forms, it seems that the scribe of the correction in P66 was using a very similar script to that used in these other manuscripts. This script is designated by Orsini and Clarysse as "Biblical majuscule" and they place both Codex Vaticanus and Codex Sinaiticus into this classification (Orsini and Clarysse, "Early New Testament Manuscripts and Their Dates," 451). According to Orsini and Clarysse, this script was used as early as the second century and they place P64+67+4 into an early stream of this style and assign a date of the late second to the early third century.

John 13.19 Correction in P66 (top) and detail from P4 (bottom)

Though the image from P4 is not as high quality, I compared similar letters from P4 with the correction at John 13:19 in P66 (highlighted in red above). There are definately close similarities between these scripts, especially with the more uniform thickness of the strokes, when compared with the corresponding thinner-thicker ductus of the fourth century codices. Therefore, purely on palaeographical grounds, this interlinear correction at John 13:19 could have occurred anywhere from the third century, into the fourth or even fifth centuries. However, the uniform thickness of the strokes in the P66 correction looks closer to the script in P4, so if this fact holds any chronological weight, then the correction is likely earlier rather than later within this 3rd to fourth century spectrum.

This correction indicates that someone was using P.Bodmer II at a date sometime after the book was produced, they noticed the omission by the copyist, and added the missing text in between the lines. It's possible that this correction occurred during one of the many rebindings that P66 underwent in its (apparently long) useful life. Nongbri has a detailed discussion of this rebinding in his treatment of the appearance of quire signatures in P66. The quire signatures could be from the orginal binding, or could possibly originate from a later rebinding. Either way, the quire signatures do not correspond with the latest binding preserved in the manuscript .According to Nongbri, quire signatures are first seen in the early fourth century so it is likely that multiple rebindings occurred on P66 in the fourth and possibly early fifth centuries (See the full discussion in Nongbri, "The Limits," 29-32).

 

P.Bodmer II detail showing reinforcement repair in the spine (highlighted in red)

P.Bodmer II detail showing reinforcement repair in the spine (highlighted in red)

 

 Figures 20 and 21 taken from Nongbri "The Limits of Palaeographic Dating of Literary Papyri," pg. 30, with the quire signatures highlighted in red.

If a third century production of the codex holds true, then the manuscript was used for about one hundred years before it was repaired sometime in the fourth century. It is possible, as Nongbri suggests, that at this time the quire signatures were placed in the manuscript and perhaps, the marginal interlinear correction at John 13:19 was made by the scribe who was facilitating the rebinding (highly speculative I admit). The manuscript continued to be used and the reinforcing strips and possibly another rebinding ocurred in the later fourth or even into the fifth century and the page margins (and the remains of the quire numbers) where trimmed. Then finally, the codex was deposited into its final location in the fifth, or perhaps (if Robinson's material is included in the same deposit) into the sixth or even seventh centuries (Nongbri, "The Limits," 25). That means that this manuscript was being used anywhere from 200 to 300 years before it was deposited into its final location (see earlier blog post here). 

One final comparison of scripts is worth noting here. P.Bodmer XIV–XV (P75) also contains a marginal comment that uses a similar "Biblical Majuscule" script as that used in the John 13:19 correction of P66. 

The marginal comments in P.Bodmer XIV-XV (P75) compared with the correction at John 13:19 in P.Bodmer II (P66) 

 

Though not an exact match, the marginal comments in P75 are very similar to the correction at John 13:19 in P66. Look especially at the two circled alphas, their shape and form is very similar, though they are not exactly the same hand because there are minor differences in the size and shape of the upsilon and the omicron, and the marginal hand in P75 has no serifs or hooks like those in the correction in P66. With that said, this indicates that these new testament codices were being used and read well into the fourth and fifth centuries. This corresponds with Houston's assessment.

"The identification of such collections, and of the manuscripts within them, provides new evidence on an old question: how long did a papyrus roll last? The evidence from our collections indicates that a usable lifetime of about 100 to 125 years was common and can reasonably be considered the norm; a small but significant number of manuscripts were still usable some 300 years after they were first created; and on rare occasions a manuscript might last, it seems, for half a millennium." (Inside Roman Libraries, p. 257)

____________________________________________

Fee, Gordon, Papyrus Bodmer II (P66): Its Textual Relationships and Scribal Characteristics (Studies and Documents, 34; Salt Lake City: University of Utah Press, 1968). 

Houston, George W. Inside Roman Libraries: Book Collections and Their Management in Antiquity (Chapel Hill: UNC Press, 2014).

Nongbri, Brent, “The Limits of Palaeographic Dating of Literary Papyri: Some Observations on the Date and Provenance of P. Bodmer II (P66),” Museum Helveticum 71 (2014): 1-35

Orsini, Pasquale and Willy Clarysse, “Early New Testament Manuscripts and Their Dates: A Critique of Theological Palaeography,” Ephemerides Theologicae Lovanienses 88 (2012): 443-74. 

Royse, James Ronald. Scribal Habits in Early Greek New Testament Papyri (NTTSD 36. Leiden: Brill, 2008).

Turner, E. G. Greek Manuscripts of the Ancient World (2nd edition, ed. P. J. Parsons. London: University of London, Institute of Classical Studies, 1987) 

Review: "God's Library," by Brent Nongbri

In recent years, the study of “Christian” manuscripts as cultural artifacts has been gaining traction in the academic world. The binding, the textual formatting, aids to the reader, and other aspects of these manuscripts are being studied by researchers in order to better understand the cultural milieus in which these documents were produced and used.

Brent Nongbri is Honorary Research Fellow at Macquarie University in Sydney, Australia, and his new book, "God’s Library," is a timely contribution to this trend of studying manuscripts as ancient artifacts. Nongbri writes, “the overall effect of my research is to raise consciousness about how messy and fragmentary our knowledge about these books is” (p. 14-15).

This review begins with a detailed summary for each chapter, for the benefit of those who have not read the book and are considering a purchase. This is followed by an evaluation of a few of the central arguments posited in the work.

The first chapter, “The Early Christian Book,” does a great job at introducing the unfamiliar reader to the materials used in making ancient books (either papyrus or animal skins, p. 24-27), along with the inks used in writing the text (p. 26), though more thorough treatments can be found in other handbooks. Where this chapter shines is in the overview of the various kinds of binding and quire gatherings used in stitching codices (p. 27-36), the various extant covers (p. 38-40), and the ancient repairing of codices (p. 40-41). Nongbri makes sure to note that the differences in codex construction, the style of stitching, the use of single or multi-quires, “do not, however, have any bearing on the dates of our surviving copies” (p. 36).

The various methods used in dating ancient manuscripts is discussed in chapter two, “The Dating Game” (p. 47-82). Nongbri introduces the reader to the methods palaeographers use to establish the dates of undated literary manuscripts and highlights several problems with the discipline (p. 49-72). The use of “Radiocarbon analysis” to determine the date of manuscripts is evaluated and several problems with the technique are discussed (p. 72-80). Methods of “Ink Analysis” are examined and the case of the “Gospel of Judas” (Codex Tchacos) is brought to bear as evidence of difficulties with the method (p. 80-82). At the end of the chapter Nongbri warns the reader that “if you see reports of dates like ‘circa 150 CE’ or ‘about the year 200’ in reference to an early Christian manuscript, you should be very suspicious” (p. 82).

In chapter three, “Finding Early Christian Books,” Nongbri recounts several interesting anecdotes surrounding the discoveries of several important ancient Christian manuscript collections; “The Coptic Codices of Hamuli” (p. 86-91), and “The Coptic Codices of Nag Hammadi” (p. 108-115). Other important or interesting manuscript find locations (if known) are also surveyed.

The Following two chapters analyze the two most famous early Christian manuscript collections; chapter four “A Discovery ‘Which Threw All Others in the Shade’: The Beatty Biblical Papyri” (p. 116-156), and, the longest chapter in the book, chapter five “An Elusive Collection: The Bodmer Papyri” (p. 157-215). Tapping into the findings of his own archival research, Nongbri attempts to correct misinformation surrounding these famous book collections.

The generally accepted dates of several of the codices in these collections are re-examined and challenged. The Beatty Codex VI (Numbers, Deuteronomy) is generally thought to be the oldest manuscript in the collection (ca. 150 CE) Nongbri (citing E. G. Turner) opts for a date range of 2nd to 3rd century, stating that “Greater precision on the basis of paleography alone is simply not possible” (p. 150). The date assigned to Beatty Codex I (P45), which has recently been re-evaluated as belonging to the 3rd century by Orsini and Clarysse, is also questioned. This is because, it is argued, “our corpus of dateable samples is taken almost entirely from rolls, and our concern here is a codex. The unstated working assumption is that the dateable rolls give a full representation of the span of time this type of writing was in use, even though the roll format was largely supplanted by the codex by the end of the fourth century” (p. 138). Though Beatty Codex II (P46) has been recently dated to 200-225 CE, Nongbri contends that “no truly compelling securely dated parallels for the script of the Pauline Epistles codex have been proposed” (p. 144).

Referring now to the Bodmer collection, it is argued “that the antiquity of P.Bodmer II [P66] has been exaggerated” (p. 199). This conclusion is supported mainly by the similarity in appearance of P.Bodmer II to P.Bodmer XX and Schøyen MS 193 (a codex believed to be originally part of the ancient collection), both in the squarish shape and general appearance of script (p. 199). Nongbri also contends for a fourth century date for P. Bodmer XIV-XV (P75) by comparing the elongated codex shape (width half the height) and construction with other fourth century codices in the Bodmer collection and the Nag Hammadi Codices (p. 199-202). Along with codicology, Nongbri appeals to the broadly similar appearance of the scripts with two securely dated letters from the fourth century, P.Herm 4 and 5 (p. 202).

Chapter six, “Excavating Christian Litter and Literature at Oxyrhynchus,” introduces the history of the late 19th and early 20th century expeditions to Oxyrhynchus, Egypt by Bernard P. Grenfell and Arthur S. Hunt (p. 216-246). Personal correspondence, contemporary news reports and magazine articles are brought to bear. The chapter lists out all the Christian papyri discovered at Oxyrhynchus according to genre and frequency of pieces found and the dates assigned to them (p. 231-233). Nongbri observes that there is a notable decrease in the number of papyri from the third to the fourth centuries, which, according to him, seems at odds with the rise in power of the Christian Emperor Constantine in the early fourth century (p. 244). He surmises that this may indicate that the dates assigned to the “New Testament” material are being artificially “pulled early” (p. 245) When the texts of undated literary Christian codices are compared with these securely dated re-used rolls, it results in an artificially early date (p. 244-245).

In the final chapter, “Fabricating a Second-Century Codex of the Four Gospels,” Nongbri recounts the history of the New Testament fragments P64+P67+P4. The story begins with the “discovery” of a Philo codex allegedly found bricked up in a wall (p. 248). The codex was thought to have been placed there in hiding during the “Diocletianic persecutions” in 292 CE (p. 249). Fragments of a papyrus codex containing Luke were found with the Philo codex, commonly known as P4 (p. 249). Because it was thought that the Lukan fragments were part of the binding cartonnage of the Philo codex, then P4 must date before (likely sometime earlier) than the late third century (p. 249). When fragments of Matthew (P64, P67) were thought to originate from the same codex, Colin H. Roberts postulated these as evidence of a second century four Gospel codex (p. 254).

Through his own archival research, Nongbri discovered that the (allegedly) securely dated Philo codex was not necessarily recovered in the wall of a house in Koptos Egypt. Rather, the information likely (Nongbri surmises) came from a local antiquities dealer after it was bought on the market and thus its provenance is unknown (p. 262). Thus, the date of the Lukan fragments (P4) are not as secure as once thought and this throws the entire second century four Gospel codex theory into doubt (p. 263-267). Based on the use of ekthesis in these Gospel fragments, Nongbri postulated that they should be dated closer to the fourth century where “similar techniques of division of the text are known to occur in manuscripts generally assigned to the fourth and fifth centuries” (p. 267).

Though Nongbri’s extreme caution towards dating manuscripts too early and with an unrealistically narrow range is appropriate in the specific cases he examines, at times his suspicion comes across as overdone. In several places this skepticism undercuts his own attempts at assigning later dates to some of the manuscripts discussed.

In the first chapter, “The Early Christian Book,” he argues that the differences in codex construction, the style of stitching, the use of single or multi-quires, “do not, however, have any bearing on the dates of our surviving copies” (p. 36). He writes that,

The variety of construction techniques of surviving codices from Egypt suggests that the rise and spread of the codex format was a drawn-out process involving both diffusion of the technology from one locality to another and independent experimentation at the local level. Refinement of construction techniques at any given time would have varied widely from place to place. (p. 36)

Yet, at several places in the book Nongbri appeals to codicology as a gauge of date (cf. p. 80). For example, when suggesting that the antiquity of P. Bodmer II (P66) has been exaggerated, Nongbri highlights the similarity of the squarish shape and the construction of P. Bodmer II with the squarish shape and similarity in binding of other fourth century codices in the collection (p. 195-199). If codicology has no “bearing on the dates” then the similarity in shape and binding between the Bodmer papyri should not be marshalled as evidence of a later date for P. Bodmer II or any other undated manuscript (see also p. 202-206, et al.).

Nongbri helpfully distills dating by palaeography down to its primary assumption that “Graphic similarity equates to temporal similarity,” and this comparison is best made against securely dated writings samples rather than against other non-securely dated examples (p. 57-58). In the second chapter, Nongbri expresses skepticism towards dating strictly by palaeography, listing out reasons that “challenge the key assumption I mentioned at the outset—that graphic similarity necessarily equates to chronological similarity” (p. 64). In a later footnote, when discussing whether palaeography is strictly a “science,” Nongbri writes, “What is more certain in terms of social history is that the modern academic discipline of palaeography emerged out of the aesthetic interests and connoisseurship of elite circles in the eighteenth and nineteenth centuries” (p. 296, note 20).

If palaeography, however, is merely a product of eighteenth and nineteenth centuries’ anachronistically applying different classifications to ancient scripts, then scholars cannot use palaeography to assign dates to undated manuscripts. It appears that Nongbri intends to bring this point home throughout the book, yet he consistently uses the graphic similarity between scripts as a criterion for assigning later dates to “Biblical” manuscripts (cf. p. 202).

With this level of contradiction and circular argumentation readers will likely come away from God’s Library scratching their heads and asking themselves, ‘Can palaeography be used to date undated manuscripts, or not? Can codicology be used to date un-dated manuscripts, or not?’

Despite these glaring problems, there is much to like about this book. The Prologue (if one reads Prologues), is well written and draws the reader into the book. The work is supported by extensive archival research that is both illuminating and, in the case of some manuscripts and ancient collections, paradigm shifting. The footnotes and bibliography are extensive which will likely make this book the standard guide for future studies in the provenance of manuscripts and collections. There are plenty of maps, diagrams, charts, and images of manuscripts that aid the reader in navigating the often times unfamiliar locations and visualizing these ancient books as artifacts.

Finally, much of the skepticism and warnings of overconfidence in dating ancient manuscripts is appropriate and warranted. Nongbri does well at highlighting issues with assigning the date of a manuscript solely on palaeography. He gives several examples from the past century of scholars attributing dates to manuscripts with little or no support from securely dated writing samples or instances of the same manuscripts being assigned widely differing dates. Christian apologists, theologians of every stripe, and historians of early Christianity should heed Nongbri’s warnings and apply an extra dose of caution and transparency when drawing conclusions or basing arguments on these early Christian books.

P75, P66, and the Useful Life of Papyrus Codices

The Summer 2016 issue of The Journal of Biblical Literature published a new piece by Brent Nongbri, a researcher at Macquarie University of Sydney, Australia: 

"Reconsidering the Place of Papyrus Bodmer XIV–XV (𝔓75) in the Textual Criticism of the New Testament." Journal of Biblical Literature 135, no. 2 (2016): 405-37.

In this essay, Nongbri argues that Papyrus Bodmer XIV–XV (P75) should be dated into the fourth century (ca. 350 CE) rather than in the first half of the third century. This new proposed date is 100-150 years later than the traditional ca. 200-250 CE date. His essay follows a similar line of argumentation as in his previous article which contended for a fourth century date for P66 as well.

“The Limits of Palaeographic Dating of Literary Papyri: Some Observations on the Date and Provenance of P. Bodmer II (P66),” Museum Helveticum 71 (2014): 1-35.

One interesting line of evidence which Nongbri brings to bear in these articles is that both codices appear to originate from the same ancient library/collection, the so-called Bodmer collection. A large portion of the collection dates into the 4th and 5th centuries. Therefore, Nongbri argues, this is evidence that both P66 and P75 date into the 4th century as well.

The Useful Life of Papyrus Codices.

I do not wish to debate the particulars of Nongbri's assessment, however, what I thought interesting was that the latest pieces included in the Bodmer collection date to the 5th century. And, if some of the additional papyri that Robinson contended were part of the assemblage as well, then the latest pieces date into the 6th or 7th centuries ("The Limits," pg. 25). What this means is that the manuscripts could not have been deposited in their location any earlier than the 5th century (probably the latter part of that century). Even if one agrees with Nongbri's 4th century date for P66 and P75, then these codices were in use for at least 100-150 years before they were discarded, add another 200 years if the collection was deposited in the 7th century. If the traditional dating holds for these codices (early third century) then we are looking at a useful life of 200-250 years, or, if the Bodmer papyri were deposited in the 7th century, 400-450 years of useful life.

In his masterful study of Roman libraries, George Houston noted the useful life of papyrus bookrolls:

"The identification of such collections, and of the manuscripts within them, provides new evidence on an old question: how long did a papyrus roll last? The evidence from our collections indicates that a usable lifetime of about 100 to 125 years was common and can reasonably be considered the norm; a small but significant number of manuscripts were still usable some 300 years after they were first created; and on rare occasions a manuscript might last, it seems, for half a millennium." (Inside Roman Libraries, p. 257)

The useful life of P66 and P75 appear to fall within the upper end of Houston's assessment. However, it is important to note that Houston's observations were specifically in regard to bookrolls and not codices (for the difference between bookrolls and codices read here). In the case of the Bodmer papyri, we have at least one instance from antiquity that confirms Houston's assessment with regard to the useful life of the papyrus codex as well.

The Marks of a Long Useful Life.
 
Nongbri notes that P75 had been rebound and repaired sometime in antiquity and bore other marks of continued use ("Reconsidering," pg. 431-432). Besides the evidence of rebinding, P75 has several interesting marginal notations that are not made by the original copyist's hand. Nongbri notes that at least two of these marginal notations use a 4th or 5th century majuscule hand ("Reconsidering," pg. 432-433).

P75 Marginal Note in Later Majuscule Hand

P75 Marginal Note in Later Majuscule Hand

Whether or not one agrees or disagrees with Nongbri's dating of P66 and P75 into the 4th century, he has teased out some very interesting details that highlight the long useful life of papyrus codices. Those who read them in antiquity obviously valued them. They repaired these books and rebound them. They studied their texts while making notes in the margin. These codices are more than just husks that carried texts, they are ancient artifacts that inform us on how early Christians used and valued these books.
(For more discussion on the useful life of papyrus books, read here, and read here)

________________________
Bibliography

Houston, George W. Inside Roman Libraries: Book Collections and Their Management in Antiquity. Chapel Hill: UNC Press, 2014.

Nongbri, Brent. "Reconsidering the Place of Papyrus Bodmer XIV–XV (𝔓75) in the Textual Criticism of the New Testament." Journal of Biblical Literature 135, no. 2 (2016): 405-37.

____________. “The Limits of Palaeographic Dating of Literary Papyri: Some Observations on the Date and Provenance of P. Bodmer II (P66),” Museum Helveticum 71 (2014): 1-35.

Greek Palaeography: Bookhands of the Imperial Peak

The previous post discussed the type of writing used for copying Greek literature, the bookhand. During the Roman imperial era, this handwriting was bilinear, that is, each letter was written between two parallel imaginary lines. This period is often referred to as the "Imperial Peak" because there is an increase in the number of ancient papyri preserved from this time. Several styles of bookhands developed during the Roman era. This abundance of writing samples has allowed scholars to study the development of these various hand writing styles. This is good news for scholars who are assigning dates to early Christian papyri. The earliest copies of the New Testament writings have been dated during this "imperial peak," from the 2nd through the 4th centuries. In regard to the Roman period, Cavallo wrote;

"[Bookhands], in the course of this period not only reach full maturity but also exhibit a typological variety that allows us to distinguish a whole series of graphic streams sometimes also visible in documents, thereby enabling us to arrive at better-founded dates." (Cavallo, "Greek and Latin Writing in the Papyri, p. 112)

During this period, there are several main types of what palaeographers refer to as “scripts,” that is, a type of handwriting used by a scribe that has common stylistic features.

Round Majuscule (Turner: Formal Round)

(In use ca. 1st to the 2nd century CE)

Characterized by the “Hawara Homer” papyrus

Biblical Majuscule

(In use ca. 2nd to the 9th or 10th century CE)

Characterized by Codex Sinaiticus and Vaticanus

Alexandrian Majuscule (Turner: Informal Round)

(In use ca. 2nd to the 8th century CE)
Characterized by PSI V 446

Severe Style (Turner: Formal Mixed)

(In use ca. 2nd to the 3rd century CE)

Characterized by the “Bacchylides” papyrus

Ogival Majuscule (Pointed Majuscule)
Developed from the "Severe Style"
(In use ca. 3rd to the 11th century CE)
Upright type characterized by P.Oxy XI 1352
Sloping type characterized by PSI X 1165

Though the development of these scripts are discernible in the papyri, palaeographers are careful not to assign too narrow a date to any particular manuscript. Turner wrote,

"For book hands a period of 50 years is the least acceptable spread of time. A palaeographer familiar with the material will refuse assent to a precise date allocated to a manuscript simply by comparison with other texts and by no other criterion." (Turner, GMAW, p. 23)

_________________
Select Bibliography

Bianconi, Daniele. "Greek Palaeography." Pages 297-305 in Comparative Oriental Manuscript Studies: An Introduction. Edited by, Alessandro Bausi, Pier Giorgio Borbone, Françoise Briquel-Chatonnet, Paola BuziJost Gippert, Caroline Macé, Marilena Maniaci, Zisis Melissakis, Laura E. Parodi, Witold Witakowski, Eugenia Sokolinski. Hamburg: Tredition, 2015.

Cavallo, Guglielmo. "Greek and Latin Writing in the Papyri." Pages 101-148 in The Oxford Handbook of Papyrology. Edited by Roger S. Bagnall. Oxford: Oxford University Press, 2009.

Orsini, Pasquale and Willy Clarysse, “Early New Testament Manuscripts and Their Dates: A Critique of Theological Palaeography,” Ephemerides Theologicae Lovanienses 88 (2012): 443-74.

Turner, E. G. Greek Manuscripts of the Ancient World. 2nd edition. Edited by P. J. Parsons. London: University of London, Institute of Classical Studies, 1987.

 

(Round Majuscule) "Hawara Homer" P.Haw 24-28

(Biblical Majuscule) Codex Sinaiticus

(Alexandrian Majuscule) PSI V 446

(Severe Style) Bacchylides Papyrus [PSI V 446]

(Upright Ogival Majuscule) P.Oxy XI 1352

(Sloping Ogival Majuscule) PSI X 1165