Early Christian Gospels: A New Study

Scott D. Charlesworth’s major study will deserve notice by anyone working on the text and transmission of the Gospels in the earliest centuries: Early Christian Gospels: Their Production and Transmission (Papyrologica Florentina, 47; Firenze: Edizioni Gonnelli, 2016). This is a wide-ranging work (developed from his PhD thesis), “thick” with data and analysis, and it would require much more space than I can give here to do justice to it. So, only a brief summary at this point, with a few preliminary comments.

The main claim is that certain formal features (or copying conventions) characterize copies of canonical Gospels in the 2^nd/3^rd centuries that were prepared for “public” (liturgical) use. Copies of these same texts prepared for private/personal usage tend to lack these formal features. The formal features that characterize copies of canonical gospels intended for public/liturgical use also distinguish these from copies of other early Christian texts, especially non-canonical gospels.

These formal features include (1) a preference for a particular codex size/shape (third-century gospel codices the same width as 2^nd century copies, but taller), (2) various “reader’s aids” such as punctuation, sense-unit demarcation, and careful copying with few ligatures, generous line-spacing, and a “bookhand” or tending in that direction.

The regularity and frequency of these formal features/conventions likely reflect early “copying centres” (connected to churches in major urban centres such as Rome). Not medieval scriptoria, but church-based centres involving only two or more copyists. There were, to be sure, local variations, for the copying conventions spread trans-locally through Christians interchanging texts and communicating with one another. But, Charlesworth insists, the evidence reflects the rapid and wide acceptance of these copying conventions applied to gospel texts that were intended for liturgical reading.

I think that Charlesworth is at least basically correct that we can see visual/physical differences between copies of Christian texts likely prepared for public reading on the one hand, and copies prepared for personal/private use on the other hand. I’ve noted an example of the latter in study of the fragment of the Gospel of John known as “P22”.[1] Also, he is correct that the extant early copies of “non-canonical” gospel texts, such as the Gospel of Mary, and the Gospel of Thomas, suggest that they were mainly prepared for individual usage, not for liturgical reading.[2]

But note that Charlesworth appears to have confined his comparisons to copies of NT gospels and extra-canonical gospel-like texts. A full substantiation of his argument would require also comparison of copies of other texts that we know enjoyed a popularity, such as Shepherd of Hermas.

Charlesworth also contends that there was a two-stage development of early Christian conventions about the nomina sacra.[3] The initial development involved a small set of words that typically had God and/or Jesus as referents: θεος, κυριος, Ιησους, Χριστος. Then, however, as further words came to be treated as nomina sacra, and it became a complicated task for copyists to judge when the referent was or wasn’t “sacred,” leading to some confusion among copyists about when to write words such as κυριος, πνευμα, and ανθρωπος as nomina sacra. So, Charlesworth postulates a second development, which was to promote a “systematic” treatment of all occurrences of κυριος and the additional words as nomina sacra irrespective of the referent. This latter step is posited to the second half of the second century. Charlesworth has made this claim for a few years now, but it will require further time and analysis to assess it. At this point, I’m not convinced that there was the coordinated action that he seems to posit. Instead, it still seems to me that what we see in the data is a general sense among early Christian copyists that certain words were to be written as nomina sacra, and a frequent confusion about exactly when to do so. But this question bears further thought.

Charlesworth also posits a clear difference between the process involved in the production and transmission of non-canonical gospel-like texts and the production and transmission of what became canonical texts. In the main, he judges, copyists of proto-canonical texts did not make major changes in the texts copied. By contrast, those who produced non-canonical gospels clearly engaged in an authorial and compositional effort, drawing upon emergent canonical gospels and freely composing new material. But, he insists, it is invalid to cite the freedom exercised in the composition of these new non-canonical texts as a basis for positing a similar freedom exercised in the copying/transmission of the canonical gospels.

Charlesworth offers an analysis of the textual variants in 2^nd/3^rd century copies of the canonical gospels (focusing on textual portions preserved in two or more early copies), contending that they show a generally high level of faithful copying. There is no evidence of a “wild” freedom in copying these texts, no indications of major insertions or deletions or re-arrangement of the text in the extant early papyri. So, he reasons, the text of the canonical gospels conveyed in our early MSS likely preserves substantially the “original” text, with only minor variations that don’t affect significantly the sense of the text. My own study of early manuscripts, and the recent study by my former PhD student, Lonnie Bell on earliest papyri of the Gospel of John, lead me to a similar judgement.[4] The early papyri exhibit variants, certainly, but these are almost entirely minor variations in word-order, tense, presence/absence of the definite article, etc.

I am pleased to see Charlesworth’s book published, knowing for a couple of years that it was in the works. It should now receive close attention from other scholars. Charlesworth is forthright and confident in his claims, and so it will be appropriate and necessary for them to be tested.[5]

[1] Larry W. Hurtado, “A Fresh Analysis of P.Oxyrhynchus 1228 (P22) As Artefact,” in Studies on the Text of the New Testament and Early Christianity, eds. Daniel M. Gurtner, Juan Hernandez Jr. and Paul Foster (Leiden: Brill, 2015), 206-16.

[2] Larry W. Hurtado, Larry W. “The Greek Fragments of the Gospel of Thomas As Artefacts: Papyrological Observations on Papyrus Oxyrhynchus 1, Papyrus Oxyrhynchus 654 and Papyrus Oxyrhynchus 655,” in Das Thomasevangelium: Entstehung – Rezeption – Theologie, eds. Jörg Frey, Enno Edzard Popkes and Jens Schröter (Berlin: Walter de Gruyter, 2008), 19-32; Christopher Tuckett, The Gospel of Mary. Early Christian Gospel Texts (Oxford: Oxford University Press, 2007).

[3] See my discussion of the nomina sacra in my book, The Earliest Christian Artifacts: Manuscripts and Christian Origins (Grand Rapids: Eerdmans, 2006), 95-134. I also discuss the early Christian preference for the codex (43-89), and several other features of early Christian manuscripts (155-89).

[4] Lonnie D. Bell, “Textual Stability and Fluidity Exhibited in the Earliest Manuscripts of John: An Analysis of the Second/Third-Century Fragments with Attention Also to the More Extensive Papyri (P45, P66, P75),” (PhD, University of Edinburgh, 2015).

[5] It is unfortunate that the book comes with such a high price, €120. But it appears in a series edited by a respected papyrologist, Rosario Pintaudi, who was also one of Charlesworth’s examiners for his PhD.