In my book, New Testament Texts on Greek Amulets from Late Antiquity, I briefly drew attention to P.Vindob. G 35894: a small seventh-eighth century parchment fragment containing parts of Revelation 10:9-10 in Greek. Because it is a non-continuous text, it is not formally considered a New Testament manuscript, namely, it is not assigned a Gregory-Aland (GA) number, the numbering system for NT manuscripts. For that reason, it has received almost no attention since its initial publication in 1982. The editor, Uwe Schmidt, observed that the remaining two letters of line 1 (and possibly those of line 8) could not be from Revelation and surmised that the text occupied this single sheet, since the back is blank (i.e., it was likely not from a codex, which contained writing on both sides of the page). As reconstructed, the text has a few variants. Here is an image of the fragment, followed by Schmidt's transcription. 

Schmidt said the purpose of the manuscript can only be speculated. He said it could be part of a patristic text in which the biblical portion was being expounded upon, though he found no match. He suggested that it could be an amulet, although there are no Greek amulets with a text of Revelation and "it is difficult to imagine a context for the amulet use of this passage." Finally, he said it could be a school exercise, but opposed this view on account of the neatness of the hand. 

So, what purpose did this little parchment serve? One might point to the angelic figure mentioned in this passage in support of an amulet designation, since divine beings were frequently invoked or merely listed in amulets. Also, the format is "miniature," typical of amulets. The lack of text on the back side is also a typical feature of amulets. On the other hand, the hand is practiced and the right margin is generous—two rare characteristics of amulets. [Although here it could be argued that the parchment was used secondarily as an amulet. See my discussion of P.Col. 11.293, a fragment from a biblical codex used secondarily as an amulet.] I side with Schmidt here in thinking that the purpose can only be speculated. There are many citations and interpretations of the Apocalypse and other apocalyptica in texts stemming from monastic settings in Egypt. Perhaps it was an amulet. Perhaps it was a little parchment carried around by a monk who liked this particular passage of scripture. Whatever the case, this is one example of many neglected Christian manuscript fragments. So, I leave it here for my readers to ponder!

The latest issue of the Bulletin of Biblical Research (25.1) features an article by NT scholar Craig A. Evans that has been the subject of much discussion on social media over the last few weeks. Evans’ article, titled “How Long Were Late Antique Books in Use? Possible Implications for New Testament Textual Criticism,” has three main arguments. The first argument, which relies on George Houston’s work on ancient libraries, is that some literary manuscripts were in use over a long period of time, sometimes for 100 years or more. The second argument, which hinges on the first, is that the autographs of the New Testament probably also remained in use (i.e., being copied, read, and circulated) for 100 years or more. The third and final argument is that, since the autographs and first copies probably continued to be copied throughout the second and on into third centuries, the New Testament text was stable and therefore did not undergo any major changes.

There are many problems with Evans’ article, but here I would like to respond to a few of the more serious ones. It should be stated at the outset that the New Testament autographs—i.e., manuscripts containing the original or authorial text—are lost to us today. Just like any other ancient text, we do not possess the manuscripts that the original authors of the New Testament wrote and we do not know what their texts looked like immediately as they left their hands (or mouths, if produced via dictation). Our earliest manuscript evidence of the New Testament is a handful of fragments, most containing only a few verses, dating from the second and third centuries. The majority of our earliest evidence for the text of the New Testament stems from the fourth and fifth centuries and is represented by major codices like Codex Sinaiticus, Codex Vaticanus, Codex Alexandrinus, and Codex Bezae. The text of the New Testament as it existed in the first and second centuries has been a point of debate for many years, primarily because we have no first century evidence, and only minimal evidence for the second century. But what must be remembered is that the autographs are lost to us today and we do not know what those texts looked like exactly.

The most controversial claim in Evans’ article is that the autographs circulated and continued to be copied throughout the second century, and even on into the third century in some cases. There is not a shred of evidence for this claim. As mentioned above, Evans bases this claim on the work of Houston. So what does Houston say? According to Houston, who gathered only a small amount of data for ancient libraries, some literary manuscripts can be shown to have had “a useful life of between one hundred and two hundred years” (251). It is apparent that some literary manuscripts were indeed kept in use for this period of time before being discarded. (It should be noted that there are some problems with respect to how Houston’s “concentrations” are established, especially in regard to the dating of papyri and the potential contamination of a collection; see 248.) But that was certainly not the normal practice. Houston himself even admits that “a considerable majority of volumes in our concentrations were not that old when they were discarded or (in the case of the Villa of the Papyri) destroyed” (250). Manuscripts were frequently retired, discarded, or destroyed for various reasons (see examples and further discussion in AnneMarie Luijendijk’s work on “Sacred Scriptures as Trash”). Multigenerational use of literary texts was an exception to the rule and so extrapolating Houston’s data to permit longevity of the New Testament autographs is highly problematic.

In the end, we have no idea how long the New Testament autographs were in use. If they were treated like other literary manuscripts of the time, they were probably discarded within a generation or so. But even if they were “kept alive” for a century or more, as Evans wants us to think, we have no evidence that they were still being copied. Thus, Evans’ argument that the autographs were “in a position to influence the form of the Greek text” in the late second and early to mid-third centuries is sheer guesswork. How could we possibly know this when we do not even possess the autographs?! Evans ultimately attempts to use this scenario of preservation and use to argue for “the textual stability of the writings that make up the Greek NT” (35). In other words, he sees his imagined autographs as filling up the gaps in our early evidence. The idea that these imagined autographs afforded textual stability or control of the early New Testament texts is completely dubious. If this were the case, then how in the world does one explain the textual diversity in the earliest manuscript tradition?

New Testament scholar Michael Kruger read Evans’ article and concluded that “this makes the gap between our copies and the autographs shrink down to a rather negligible size.” In reality, however, not a single thing has changed. Evans has not discovered new evidence: he has invented it. There are no new links between the autographs and the earliest manuscript copies despite what Evans wants us to think. The gaps have not shrunk down to a rather negligible size, as Kruger claims. The autographs of the New Testament are lost, and we have no idea what happened to them. Thus, Evans’ arguments about the longevity of the autographs and their influence on the manuscript tradition are built wholly on multiple, untenable assumptions.

Like Evans, many scholars continue to overemphasize the “early” in order to argue for textual stability, but this method is flawed. It is wrong to assume that “earlier” manuscripts always contain better readings and that late manuscripts always contain bad readings. Indeed, readings in later witnesses have been found to have early support. As J.K. Elliott has rightly warned, “to emphasize their [i.e., the papyri] early dates is deceptive. The age of a manuscript is of no significance when assessing textual variation, unless we know how many stages there were between the autograph and that copy and also what changes were made at each of the intervening stages. No one has such information” (223).

Evans attempts to strengthen his claim that the early text of the New Testament was stable by turning to the “Gnostic manuscripts.” According to Evans, the “Gnostic” writings were less numerous and less sacred than the New Testament texts, were read and studied in private, and were not taken as seriously as New Testament texts (“the NT writings were taken more seriously by their readers and copyists, with the Gnostic writings—probably read and studied in private—seen more or less as ‘interpretations’ of the dominical and apostolic traditions” [36]). These points, he suggests, “seem to show significant instability in the Gnostic manuscripts—in marked contrast to the NT manuscripts, whose text is considerably more stable” (36). We can probably all agree on the first point (i.e., the Nag Hammadi texts were less numerous than New Testament texts). But how in the world can we know that the Nag Hammadi texts were considered less sacred? And the assumption that readers took the text of the New Testament more “seriously” than the Nag Hammadi texts is absurd. In short, Evans appeals to a canonical-bias approach here: he undercuts the non-canonical texts by showing that they were unstable, unpopular, read in secret, and so on, in order to argue for the superiority and textual stability of the canonical texts.

All in all, I—and a whole slew of other scholars—am baffled as to how this article, full of faulty assumptions and claims, came to see the light of day. Its publication in the Bulletin of Biblical Research demonstrates that this journal’s editorial and peer-review standards seriously need to be reevaluated.


Update: See contributions to this discussion by my astute colleagues Brent Nongbri, Malcolm Choat, J.K. Elliott, Gregg Schwendner and others in the comments section below.  

Elliot, J.K. “The Early Text of the Catholic Epistles,” in The Early Text of the New Testament (ed. Charles E. Hill and Michael J. Kruger; Oxford: Oxford University Press, 2012), 204-224.

Evans, Craig A. “How Long Were Late Antique Books in Use? Possible Implications for New Testament Textual Criticism,” Bulletin of Biblical Research 25 (2015), 23-37.

Houston, George W. “Papyrological Evidence for Book Collections and Libraries in the Roman Empire,” in Ancient Literacies: The Culture of Reading in Greece and Rome (ed. William A. Johnson and Holt N. Parker; Oxford: Oxford University Press, 2009), 233-267. 

Luijendijk, AnneMarie. “Sacred Scriptures as Trash: Biblical Papyri from Oxyrhynchus,” VC 64 (2010): 217-254.

Ed. princ. Franco Maltomini, “340. Amuleto con NT Ev. Jo. 1, 1-11,” in Kölner Papyri (P. Köln), vol. 8 (eds. Michael Gronewald, Klaus Maresch, and Cornelia Römer; Opladen: Westdeutscher Verlag, 1997), 82-95.

P.Köln 8.340, a long amulet containing both text and images, was designed as a request for healing and protection. It begins by appealing to a lengthy passage of scripture (John 1:1-11), followed by an invocation of the name of God, requesting that he send his angel to chase away sickness, evil spirits, the evil eye, and “every snare of humanity.” I am currently working on this amulet, which has a number of interesting features. Here is the text, following the NT portion, in translation (my own):

"I call upon you God, and Mary the God-bearer, Father of our Lord and Savior Jesus Christ, that you might send your angel who presides over the healing of those who wear this adjuration [amulet] and implore him to chase away each and every illness and infirmity, every unclean spirit, every evil eye, every snare of humanity. I banish you by the glorious name of the Lord forever and ever. Amen, amen, amen.”    

On the backside—in this case, we cannot speak of recto and verso because overlapping patches of papyrus strips preclude such identification—there are two drawn figures, both standing, depicted as praying in the orantes position, i.e., with their hands raised. In this post, I would like to correct the interpretation about one of these figures by the editor, F. Maltomini. According to Maltomini, there is a face superimposed onto the chest of the second standing figure.[1] He describes it as lacking hair, eyebrows, eyeballs, mouth, chin, and neck. The eye sockets are described as tiny and round and the nose as being constructed by a line beginning at the top part of the forehead extending down the bottom of the face and finally curving off to the right. Maltomini wrestles with the identification of this “face,” and concludes by suggesting that it is “probably the person for whose healing the two stand praying their prayers.”[2] 

The problems associated with the identification of this superimposed “face,” however, can be easily resolved: what Maltomini describes as a “face” is clearly, in fact, an image of a woman’s breasts. This would explain, then, why this “face” lacks hair, eyeballs, eyebrows, mouth, and chin, and why the "nose" is represented by a long curved line. This identification is further secured by the fact that the standing figure has long, wavy hair; even the editor admits that this must be a female figure on this basis. The breasts are somewhat similar in appearance to the breasts depicted in another Christian amulet, P.Oxy. 8.1077, but are drawn at more of an angle.

Does the inclusion of a female figure suggest that the owner of this amulet was a woman? Perhaps it does, although it is difficult to say who the first (presumably male) figure might be and his relation to the female figure. In my study alone I have see at least two other amulets that were clearly owned by women (P.Oxy. 8.1077 and P.Oxy. 8.1151). Nevertheless, the participial phrase τὸν φοροῦντα ("the one who bears" [the amulet]) in ll. 41-42 of this amulet seems to preclude the possibility of a female owner of the amulet, since it is masculine. 

One note on the NT text. If the owner of our amulet purchased it from a ritual specialist (i.e., a church leader), then this may mean that the text was copied from an actual manuscript, although we have no way of proving this, of course. Alternatively, since the Gospel of John was apparently popular in Egypt—for example, a high number of manuscripts of John were discovered at Oxyrhynchus—its text may have been part of the oral culture of the Christian community in which this amulet was produced and used. Either way, P.Köln 8.340 contributes to our knowledge of Egyptian Christianity in more ways than one and it, like many amulets, deserves the attention of early Christian scholars.

[1] Maltomini’s full description of this figure runs as follows: “Al di sotto di questa figura è rappresentato un orante. Il viso, appena abbozzato, si sovrappone a parte del petto della figura precedente. La linea del contorno non appare chiusa in alto sulla testa; assenti i capelli; gli occhi sono piccoli e rotondi, senza pupille e senza sopracciglia; il naso è constituito da una lunga linea che si inizia nella parte alta della fronte, scende dapprima verticale per poi piegare verso destra. Bocca, mento e parte del collo sono scomparsi in una lacuna. Il tronco è rettangolare; di alcune linee irregolari che vi appaiono all'interno non so ravvisare il significato preciso (panneggio?). Le braccia sono sollevate nel gesto della preghiera, più distese di quelle del primo orante, e vengono ad incorniciare la figura centrale. Non si distinguono gli arti inferiori” (Maltomini, “340,” 95).

[2] My translation of “…probabilmente la persona per la cui guarigione…i due oranti levano la loro preghiera,” (Maltomini, “340,” 95).