Mark 16 in Codex Bobbiensis: New Elements on the Shorter Ending.

1 Introduction

This eTalk highlights main elements of a 2021 Zeitschrift für Neues Testament article published in German by Claire Clivaz, PI of the Swiss National Science Foundation MARK16 project. This article has been graciously translated into German by Prof. Dr. Manuel Vogel. It is in open access with the kind permission of the publisher and is based on an online invited lecture given at Duke University (USA), on 16 October 2020.
To begin, let’s first remember that the hypothesis that Mark originally ended at 16:8 has dominated New Testament scholarship for several decades, even to the present day, as this 2019 article about the state of the art on Mk 16 demonstrates.
But other periods in the history of exegesis have known lively debate on the topic, a point raised by Kenneth W. Clarke’s 1965 SBL presidential address. In this lecture he noted that the 1946 Revised Standard Version reintroduced Mk 16:9-20, justifying this change in view of the history of reading by pointing to the attestations of the long ending in Justin, Irenaeus, and Tatian, as well as evidence in the Latin, Syriac, and Coptic traditions.
The history of Mark’s endings is a long one, and the SNSF project MARK16, which explores the manuscripts themselves, whether they are well known or never read before, has highlighted a long reading history with continuous debates. Among these debates, the so-called “shorter ending”, or conclusio brevior, has to be highlighted as an element that has been neglected to date. Its most ancient attestation stands in the Latin Codex Bobbiensis, the manuscripts at the center of this eTalk. I will first offer some general context on the manuscripts data on Mk 16, before turning to focus on Codex Bobbiensis and its specific version of the shorter ending.
First of all, we have several cases of folios accidentally lost, of liturgical material, or folios replaced at the end of Mark. Manuscripts that end at Mk 16:8 but only due to loss of material include: GA 1420, GA 2386, arb 2 or the Ethiopic manuscript Chester Beatty W 912. Those that end in Mk 16:8 that have liturgical material include: Ethiopic EMML 2868 and 3875. Manuscripts that have folios replaced at the end of Mark include GA 05 (Codex Bezae), in the fourth-century CE Latin Codex a (VL 3 or Vercellensis), and in the fifth-century CE Latin Codex n (VL 16).
An end at Mk 16:8 is present for the Greek manuscripts in GA 01, GA 03, GA 304. Some examples in other languages are the Coptic sa 1, the Armenian arm 252, and the Coptic Amulet sa 393var.
The case of GA 304 is particularly interesting: Mina Monier, a post-doc on the MARK 16 project, has published the first article ever devoted to it. This manuscript indicates a clear desire to end the text at 16:8 while simultaneously showing signs of an ongoing discussion around the final sentence, which was written and then at some point erased: Ὥσπερ ξένοι χαίρουσι ἰδεῖ[ν πατρίδ]α, οὕτω καὶ ὁ γράφοντoς βιβλίο[υ τέλο]ς ; “As the travellers rejoice upon reaching their homeland, likewise the scribe is upon the end of this book”.
Claire Clivaz and Mina Monier have continued to demonstrate in articles the uninterrupted discussion of Mark’s endings. Consequently, it is not possible to simply claim with Croy that “acceptance of the genuineness of the Long Ending prevailed for centuries”. Rather, Mark’s endings have been discussed for centuries.
But within this ongoing discussion, an element has clearly been underestimated in scholarship: the so-called shorter ending, or conclusio brevior, of the Gospel according to Mark. Codex Bobbiensis – also called k, or VL 1 or G.VII.15 – has the oldest attestation of the shorter ending but is not the only evidence for this Markan ending. Moreover, this Latin version is slightly different from the Greek one: as Matthew Larsen summarized it, “Codex Bobbiensis (k) may also be regarded as yet another ending”.
Mk 16:8 followed by the shorter ending without a break can also be found in GA 044 and GA 579, and in 131 Ethiopic manuscripts. The number of Ethiopic manuscripts with the shorter ending was increased in the second edition of Bruce Metzger’s article, but this point is seldom noticed. Indeed, he added a list of Ethiopic manuscripts produced by Dr. William F. Macomber.
GA 083, 099, 019, l1602, syhmg, sa 9 also appear to have been designed to conclude with the shorter ending. Two new Greek testimonies on this point have been found by Mina Monier in the catenae of GA 1422 and GA 2937.
Evidence for the shorter ending exist in at least six ancient languages and in about 151 manuscripts. You have the list on the slide, as I published it in 2021.

2. An overview of existing scholarship: Towards a reconsideration of Codex Bobbiensis (VL 1)

Counted as number 465 in the fourth volume of the Codices Latini Antiquiores, G.VII.15 or Codex Bobbiensis has “uncertain origin”, according to Elias Avery Lowe. He suggested nevertheless that it was produced in Africa according to “two considerations: the text of k is nearest to the Gospel text used by Cyprian, an African writer, and the peculiar type of uncial in k has its nearest parallels in two fifth-century MSS of Cyprian”.
Convinced of the importance of this manuscript for New Testament textual criticism (NTTC) digital scholarship, the MARK16 project has asked the Institut für Neutestamentliche Textforschung (INTF) to insert Codex Bobbiensis in the digital reference Liste and to give to it a Doc ID number in the New Testament Virtual Manuscript Room (NTVMR). The chosen name by the INTF is VL 1, Vetus Latina 1, according to the collective name given to the ancient Latin versions of the Bible by Bonifatius Fischer. I have decided to follow this choice of the INTF, and to designate k mainly as VL 1.
The MARK16 project is working daily with the NTVMR, building our own manuscript room focused on Mk 16, including images and a transcription of VL 1 f. 40r, f.40v, and f.41r, with our gratitude to the National Library of the University of Torino.
This naming situates VL 1 clearly within NTTC research and underlines a kind of “retour en grâce” for this codex, which is often considered to be carelessly copied, as stated in an exemplary way by Bridget Gilfillan Upton, who says that its “language itself is impenetrable almost to the point of incomprehension”. This general feeling is often intensified by the fact that the codex is not well known; see for example Jean-Claude Halewyck who did not consult it to establish his 2013 edition of the Vetus Latina, convinced that the manuscript has become totally unreadable. But scholars like Wordsworth, Parker, and Lupieri have devoted attention required to it in an effort to do justice to this manuscript and its particularities.
As examples of positive opinions about Codex Bobbiensis, let’s quote Bruce Metzger in 1977: “It is the most important, as regards text, of all the Old Latin copies, being undoubtedly the oldest existing representative of the African type. […] The scribe, though committing many blunders in writing, was not uneducated, for he writes with a firm and practiced hand.”
We can also quote here the important study by Hoogterp in 1930, which includes this convincing hypothesis: Codex Bobbiensis is a direct copy of “a very ancient archetype in cursive writing, perhaps difficult to decipher for a copyist who had only some vague notions of correct Latin.” This laborious copy of a third-century cursive could explain several mistakes present in VL 1.
Codex Bobbiensis has never been forgotten by Classics scholars or New Testament textual critics, notably represented by an important article by David Parker in 1991, republished in 2009. Nonetheless, Bobbiensis has been largely ignored in New Testament exegesis in the last decades until the early 21st century.
Attention to Bobbiensis has reappeared at the forefront of New Testament exegesis, and this new step in the state of the art is extensively presented in the ZNT article. One can notably mention the unpublished PhD of Jayhoon Yang (2004), an article of Camille Focant (2006), and the monograph of Bridget Gilfillan Upton (2006). Codex Bobbiensis has now been the recipient of new interest in New Testament exegesis and is more regularly mentioned, as it is in Kara J. Lyons-Pardue’s 2020 monograph.

3. Qui cum puero erant: a mistake or not in VL 1?

Let’s turn now our attention toward the most surprising point of the shorter ending in VL 1: the clearly written mention of qui cum puero erant, “those who were with the boy”, instead of qui cum Petro erant, “those who were with Peter”. In the concerned folio, one reads absolutely clearly qui cum puero erant. Legibility is not at stake. In the last decade, Haelewyck (2013) and Larsen (2021) have considered this reading to be a pure mistake, whereas Hugh Houghton has chosen to leave the point open in his 2016 edition and translation: “But those who were also with the boy [for Petro, Peter?] told in brief everything which they had been instructed. After this, Jesus himself appeared too and sent the holy and unchanging <message> of eternal salvation through them from the east all the way right to the east [west?]. Amen”.
You can find a complete overview of scholarship on this point in my article. It should be underlined that, when the scholars correct puero, they usually also change et qui to eis qui cum Petro erant to find a wording similar to the Greek. But other scholars have kept et qui cum puero erant, as it is indeed clearly readable. Secondly, according to Hoogterp, the manuscript “a été copieusement corrigé”, that is, largely corrected. It is consequently highly surprising that the correctors have kept puero in f. 41r. One has to seriously consider this point because numerous other points, also in this folio, have been corrected.
Four questions result from the state-of-the-art overview that have to be answered in order to establish an edition and translation of the shorter ending in VL1: 1) Can one find a plausible framework for understanding why the correctors have kept puero? 2) If yes, was this framework already present at the time of the copy of VL 1, or is puero a pure misreading from a cursive exemplar (Hoogterp)? 3) Is it possible to affirm that puero was not present in the manuscript copied by VL 1? 4) Last, but not least, if puero is a cursive copy mistake, is it mandatory to also correct et qui to eis qui?
1) Can one find a plausible framework for understanding why the correctors have kept puero? Let’s distinguish the time of the writing of VL1 and its reception and correction. Considering the manuscript’s reception and correction periods, the polymorphic Christology could explain why et qui cum puero erant has not been corrected, considering that other points have been corrected on f.41r. Numerous texts like the Acts of John – widespread in Early Christianity – attest to the existence of such a Christology, whereas Paul Foster considers the expression ἐν ἑτέρᾳ μορφῇ in Mk 16:12 as proof of the “growing importance of the polymorphic Christ during the second century and later”. Moreover, Eric Junod has demonstrated that polymorphic traditions in Egypt are often related to the sun god Horus. It is coherent with the spectators’ viewpoint in Mk 15:35 in VL 1: “et quidam eorum qui aderant cum audissent aiebat ‘helion uocat’”; “and one of those who were there, while they were listening, was telling ‘he calls Helion’”.
2) If yes, was this framework already present at the time of the copy of VL 1, or is puero a pure misreading from a cursive exemplar (Hoogterp)? In my opinion, it is impossible to decide whether the copyist was convinced by such a cultural framework to the point that he/she read and wrote puero or whether he/she made a pure misreading from a cursive exemplar, as plausibly argued by Hoogterp. Both possibilities can also be mixed up: the first scribe of VL 1 could have been inclined to read puero in the cursive exemplar because of his/her knowledge of Christological polymorphic traditions. The answer to this second question will thus remain open but is not decisive as such for the edition.
3) Is it possible to affirm that puero was not present in the manuscript copied by VL 1? Three strong arguments allow to answer “yes”. First, even if Mark 16 presents several variants, no other element in this chapter could lead to reading the presence of a Christological polymorphic tradition. Second, in all other versions of the shorter ending that I have been able to read so far, “Peter” is clearly written, without alternative. Third, and finally, the VL1 shorter ending serves, at the narrative level, as a final narrative lock (verrou narratif): everything is exposed breuiter, in a shorter way. There is surely no space here for stories of unusual appearances or for further teachings.
4) Last, but not least, if puero is a cursive copy mistake, is it mandatory to also correct et qui to eis qui? To summarize the answer to this crucial question, two main arguments allow us to maintain et qui with Petro in the edition of the text. First, other linguistic versions also attest that Peter’s companions can be the subject of the reporting. It is the case of the Syriac version, in the Chaldean 25 manuscript, whose shorter ending is translated in this way by Samer Yohanna: “‘It is given somewhere and these’: all these things ordered to the household of Peter we have reported briefly. Afterwards Jesus himself, through them, sent forth from east to west the sacred and imperishable proclamation of eternal salvation. Am[en].” This reporting by Peter’s people stands also in Ethiopic manuscripts, as Metzger has underlined.
You have an example here in Eth 2, visible on our VRE manuscript room, transcribed and translated by Damien Labadie, and encoded by Mina Monier. The shorter ending starts with “when they had finished saying all that he had commanded”. The reporting subject is “they”, a male pronoun, so not the women.
Second, the double return of the women to Galilee on one side, and of Peter and his companions on the other side, is found in another Gospel ending, the Gospel of Peter 54 and 58, as underlined by Eric Junod. The women come “back home” (Gos. Pet. 54) without having received an order of transmission from the young man at the tomb; their silence is not mentioned in Gos. Pet. 57, just like in VL 1 Mk 16:8b.
In summary if we consider the shorter ending in languages other than Greek, we have attestations that Peter’s companions are reporting. Consequently, I keep et qui in the edition of the text as an echo of this alternative tradition, a choice also made previously by H. A. G. Houghton.
This point and others presented in the published article have served as basis for the choices of the transcription that Mina Monier and I produced, based on Wordsworth et al.’s transcription, and benefitting from H.A.G. Houghton’s suggestions.
In the article, I have provided a critical edition, with a complete critical apparatus and notes, and a French translation.
It seems obvious to me that interest in Codex Bobbiensis and the shorter ending won’t stop anytime soon. To be able to examine these three folios online is a benefit for all scholars in the field. I want to conclude by warmly thanking the Library of the University of Torino for its collaboration with MARK16. Nothing can replace the face-to-face meeting with a venerable manuscript.
References https: