Why assimilation theories of Bezae’s Greek text are compelling

In an earlier post, How to validate a theory of Bezae’s text, I suggested a fourfold classification of the nine theories of the so-called “Western” text enumerated by B. M. Metzger in his Textual Commentary. [1] The four categories are:

  1. Multiple editions
  2. An initial “Western” text
  3. A secondary “Western” text
  4. Assimilation to either the Latin or Syriac versions

(Of course, reference to the “Western” text is somewhat problematic given the existence of multiple “long texts” of Acts. Therefore I will consider these theories from the point of view of the single extant text of Codex Bezae.)

It is evident that the first three classes of theories focus on the timing of the development of the “Western” tradition with respect to the mainstream tradition — whether it came before, after, or roughly contemporaneously with the mainstream. Theories of assimilation also presume that the “Western” text is secondary, but go further in attributing the development of this text form to a specific source of influence, namely, a desire to conform the Greek text to a more familiar versional text.

Now the consensus is not favorable to theories that Bezae’s Greek text reflects systematic assimilation to a version, in particular, to the Latin version. As Metzger plainly observes:

“The theory finds little or no support among present-day scholars.” [2]

It seems though that most theorists since the eighteenth century have considered the question primarily in terms of the interaction of Bezae’s columns, where it has been shown that assimilation to the Latin column cannot explain the development of the present Greek column. Framed in these terms Latin assimilation is easy to dismiss.

But as I have argued in “Has Latin influence on Bezae’s Greek text been disproved?” and “Against reason? Bezae’s Greek text and the possibility of Latin influence,” interaction between the columns is just one aspect of the question of Latin influence. A second, more fundamental aspect is whether Bezae’s Greek text has been corrected to one or more Old Latin exemplars besides its Latin column.

Once this bigger picture is considered, the theory of assimilation to the Old Latin begins to offer what other theories for the most part lack, namely, a specific, documented historical context and, more importantly, a compelling motive for the effort and expense of producing a Greek text form that so often mirrors the Old Latin version. This motive was simply the belief among certain of the participants that the Old Latin version was less corrupt than the Greek. Thus, in his Commentary on Romans 5:14, we find Ambrosiaster arguing that the text form found in his Old Latin manuscript, lacking the negative particle, is in fact the correct reading, while the Greek manuscripts that have the particle are corrupt.

Nevertheless, it has often been considered next to impossible that the ancient producers of a Greek text would borrow from a Latin version. H. Marsh argues:

“I have myself collated the two first chapters of St. Mark … and have found that in most of the readings, in which the Codex Bezae differs from all the Greek manuscripts, it agrees with some one of those Latin versions. But shall we therefore conclude that those readings were actually borrowed from a Latin version, and translated into Greek? It is at least as possible that they might have had their origin in the Greek as in the Latin, and this very possibility is sufficient to defeat the whole of Wetstein’s hypothesis [that Bezae borrows Old Latin readings].” [3]

Marsh concludes:

“there is no reason whatsoever for ascribing any reading of a Greek manuscript to the influence of the Latin, unless it can be proved that it could not have taken its rise in the Greek, and that it might easily have originated in the Latin” [4]

Although Marsh acknowledges the abundant evidence and correctly infers what the evidence seems to imply, he simply assumes that no one would ever want to correct a Greek MS to a Latin MS. Perhaps more tellingly, he speaks of readings being borrowed and translated or having an origin as though these events took place in a vacuum. Of course, what is missing from Marsh’s account is any attempt to recover the perspective of the human participants or even to acknowledge that human actors exist.

Yet going back to Ambrosiaster’s argument, it is obvious that a corrected Greek MS without μη would show not a trace of having been inspired by the Latin reading. The absence of the particle really leaves no definitive argument as to why it might be lacking. Certainly, Latin influence is not the most compelling explanation for a missing particle. So by Marsh’s rule we have no reason to suspect Latin influence. Yet we have a documented case of an unambiguous motive to drop this very Greek particle in accordance with the Old Latin variant. The motive of course is that the Old Latin version was regarded as preserving the true text, but the Greek was seen as corrupt.

But can there be any more compelling motive for the correction of a Greek MS than the belief that it was simply wrong?

It turns out then that the compelling aspect of assimilation theories is the human aspect and, in particular, the practically limitless human capacity to insist that what is familiar must be correct. Clearly, human participants are dangerous to text-critical theories based on pure reason. Once they are allowed into the picture, assimilation theories that were once judged impossible by the standards of criticism are suddenly not so improbable and in the proper context even compelling.

[1] B. M. Metzger, Textual Commentary on the Greek New Testament (2d ed.; Stuttgart: Deutsche Bibelgesellschaft, 1994) 223-232.

[2] Metzger, Textual Commentary, 231.

[3] H. Marsh, “Notes” in J. D. Michaelis, Introduction to the New Testament (trans. H. Marsh; 1802) 2/2:680-681.

[4] Marsh, “Notes,” 2/2:683. J. S. Semler likewise considers it “against reason” that a Greek MS would be corrected to a Latin copy. J. S. Semler, Apparatus ad liberalem Novi Testamenti interpretationem (1767) 44.

11 thoughts on “Why assimilation theories of Bezae’s Greek text are compelling

  1. I must point out that, if proven, this would not be the only case in which a Greek text has been corrected in accordance with a Latin standard—the case of 1 John 5:7, which almost undoubtedly began as a Latin addition comes to mind. Even clearer is the example of Revelation 22.19. Every Greek manuscript, with the exception of a few marginal readings, and a copy made from Erasmus’s printed text, reads, απο του ξυλου της ζωης, a reading that is contextually vastly superior as well. There is a Latin variant, followed by Erasmus (since he back translated these verse from Latin) that replaced ligno with libro. Scrivener in his 1881 edition of the TR, reverse edited to match the decisions of the KJV translators, followed Erasmus’s text. As crazy as it may seem, I personally know more than one person who would insist, to the point of making it a “test of orthodoxy” against any and all evidence to the contrary, that Erasmus and the Latin were correct and every single Greek manuscript is wrong. The spirit of Ambrosiaster is not an isolated phenomenon in church history. One might also think of the “Old Believers” movement in Russia as another example of a similar thought pattern.

  2. Thank you, Peter. These examples are certainly indicative of the human dimension I wanted to bring out, as you aptly call it, “the spirit of Ambrosiaster.” The Comma Iohanneum is a superb illustration of the conflicting attitudes towards the Latin version in ecclesiastical versus critical circles. Thanks for pointing out this classic example!

    • Have you read Amy Donaldson’s dissertation on “Explicit References…”? She has quite a fascinating discussion of Ambrosiaster.

      • I just re-read the section you mention. Yes, Dr. Donaldson’s remarks are highly relevant to the discussion on Ambrosiaster’s dismissive attitude towards the Greek tradition. She mentions some additional examples that I’ll need to look into now. Thanks for the pointer!

  3. (I couldn’t figure how to post today — July 13 — so I’m jumping back to this post, just to ask a question.)

    Have you read Chase’s book proposing Syriac influence on D’s text?

  4. On many of your posts, including your most recent one, there isn’t a comment button, so I am responding here. Have you read Peter Williams essay in “The Early Text of the New Testament?” Some of his discussion would be relevant to the point you are making. Tying in what you said in this post about the “human element” with your thoughts in the other one: Suppose that a certain individual, with some level of familiarity with the practice in bilingual synagogues of giving oral, often interpretive translations, performed such a service for a particular congregation or group of congregations. This interpretive translation was written down and many sections were memorized by several congregants who later became scribes of Greek texts. Given the sort of attitude that Ambrosiaster demonstrated, is it not possible that what started out as a targum, perhaps by someone who was originally from Palestine, ended up influencing the Greek text produced by scribes whose mental furniture had been furnished in this way?

    • Hi Peter,

      Thanks for pointing out the missing comments area! It appears I need to explicitly allow comments when I change a saved page to a post.

      Also, I apologize for my delayed response. I was on a 3-day camping trip with my family to Deception Pass SP. I realize (now) that I ought to have left a notice on the blog, but was overly-optimistic about my ability to maintain an acceptable degree of responsiveness! Thanks for your patience!

      Yes, I have read Peter Williams’ essay in “The Early Text of the New Testament?” Certainly we need to be cautious about attributing readings to Syriac (or for that matter to Latin) influence that have alternative explanations in Greek. Nevertheless, it does not seem possible to account for all of the apparent Semitic parallels in Bezae with explanations in Greek.

      What you describe certainly works as a model for the Targums, though its applicability to the gospels is unclear given our limited knowledge of Jewish-Christian context within which such a process is suggested to have occurred. Unless Bezae’s text is extremely early, it seems difficult to appeal to an oral tradition as a significant source of its readings. And we still must explain why this oral tradition was relevant to Bezae’s producers in ca. 400 to account for the codes. Our late examples of interpretation in a Christian context do not seem adequate as a model for introducing new readings. When we collate Bezae with the mainstream text, it seems clear its readings in general offer very little of the consistent interpretation we’d expect in a Targum. Rather any interpretive element in Bezae seems at best sporadic and incidental to other concerns of its producers. But I agree the “human element” requires us to look at versional interaction as a source of variants. In this respect, significant epexegetical variants (that are not necessarily interpretive) from a versional source should not surprise us.

      In my view, a layered approach is more helpful in addressing the problems noted by Williams, since if there really is Aramaic or Syriac influence, we’d expect readings reflecting this influence to group together in the same layer. In fact, whether or not there is a Syriac layer ought to provide a good indication as to how prevalent distinctive Syriac influence might be. My sense is that there is not a distinctive Syriac layer, but rather a layer with primarily Latin but also some Syriac influence.

      Finally, it has been noted that Bezae’s Greek text sometimes betrays a limited mastery of Greek. Since knowledge of Greek would have been more prevalent in the East, it would not be surprising to find that Bezae’s Greek text had been translated by an individual whose native tongue was Aramaic or Syriac, which seems to me a more promising explanation for the “Syricisms” in Bezae’s text.


      • I’m not thinking of any direct influence of oral tradition or Syriac. What I’m thinking of is that the Latin translation tradition, which started very early indeed, partook of some of those characteristics but the text that began as a relatively free translation became hallowed by use. Centuries later the Greek text was altered to match the traditional Latin, carrying over some of those influences into the text of Beza and others. No idea if that is what happened but some of the things you have said seem to point in that direction.

        • ok, I think seeing “oral tradition” threw me off, since this interpretive aspect of the Targum literature I don’t think is consistent with Bezae’s more mundane pattern of variation.

          The date of the Latin versions is an interesting question that I have not touched on in the blog. I don’t think we can assume there was a Latin version before ca. 230. (This date comes from B. Fischer’s 1972 essay “Das Neue Testament in lateinischer Sprache” and is based on the relationship of k and Cyprian.)

          Regardless of the date of the Latin version, there is certainly the possibility that features of Bezae’s text entered its tradition indirectly through other primary influences. (I mentioned this in https://peterlorenz.me/2016/06/06/date-of-bezaes-text/). To answer this question, we would need to examine Bezae’s readings by layer to determine the layer in which these Syriac influences occur. Of course, we also need to agree on defining Syriac influence. William’s article shows how tricky this can be.

          But it seems we should keep open the possibility that a Greek text might be adjusted to conform with a version whose readings have become standard within a given tradition.

  5. James,

    Thanks for pointing out the missing comments area, yet going out of your way to post a comment anyway!

    (It turns out several of my recent posts began life as “pages” rather than “posts.” By default comments appear to be disabled for pages and maintain this setting when I change the post-type to “post.” I’ll need to check the “comments” box when I change the post-type or figure out how to change the default behavior.)

    To address your question, yes, I am familiar with both of Chase’s books on Syriac influence on Bezae: The Old Syriac Element in the [Acts] text of the Codex Bezae (1893) and The Syro-Latin Text of the Gospels (1895). At an earlier point in my research, I considered the Syriac / Aramaic element more compelling as a fundamental characteristic of Bezae’s text, largely on the strength of Chase’s examples. But Syriac influence does not seem capable of uniquely explaining a high proportion of variants (though given a sufficiently distinctive set of a parallels, outright Syricisms should not be necessary in themselves to demonstrate Syriac influence).


What do you think?