Wednesday, July 12, 2006

Not a Marcionite Manuscript


In an earlier post (here) I outlined the argument of Claire Clivaz that P69 (P. Oxy 2383) should be regarded 'as a witness to a Marcionite edition of Luke's Gospel' (C. Clivaz, 'The Angel and the Sweat Like "Drops of Blood" (Lk 22:43-44): P69 and f13' HTR 98 (2005), 419-440, citation from p. 420).

Here are some critical reflections on her argument.

Firstly, we should admit that the argument that P69 represents a Marcionite version of Luke (stated on p. 420; suggested on p. 429, 432, restated in conclusion, p. 439), is presented by Clivaz as a suggestion (p. 429), a proposal which offers a plausible fit for the peculiar nature of the text represented in P69. There is an appropriate hesitancy in her presentation of this argument.

Secondly, we need to note that by the nature of the evidence Clivaz’s argument is both difficult to prove and difficult to falsify. Marcion’s text of Luke cannot be determined with confidence for several reasons: no text survives; church fathers were not interested in providing Marcion’s text, but in refuting Marcion from his own text where possible. The primary characteristics of Marcion’s text are determined mostly be omissions and absences (minuses). Clivaz has to make an argument for Marcion’s likely text of this section of Luke and then (lo and behold) identify agreement in this minus between P69 and Marcion. So in effect it is a very complicated argument from silence.

Thirdly, perhaps simply stating the previous point a bit more strongly, there is no actual positive ancient evidence for ANY connection between Marcion’s text and P69.

Fourthly, P69 generally exhibits a pretty free or paraphrastic text (with numerous singular readings, numerous agreements with D although clearly independent of that type). Although Marcion does seem to have paraphrase occasionally as well as omitted material, there is no suggestion of a connection between Marcion’s text and the D-type text elsewhere (I stand to be corrected on this point). Focusing on two aspects of P69 without considering the wider evidence of P69’s text may have led Clivaz astray (ironically enough considering the amount of space given in pp. 425-426 to criticising others who have failed to reckon sufficiently with this manuscript).

Fifthly, in relation to P69’s text of Luke 22.61 Clivaz takes it as significant that Jesus is not looking at Peter, but that Peter is looking (p. 431). This she takes as avoiding the idea that Jesus paid attention to Peter, which fits with the supposedly Marcionite idea that Peter’s status would be diminished. Each of these steps are problematic.

  1. P69 doesn’t contain the required verb ‘look’ in its extant text. It does contain the last four letters of STRAFEIS followed by O P and traces which are plausibly reconstructed as PETR[OS, so there is clearly an implication that Peter turns, but no explicit indication that he looked. Clivaz appeals to the agreement of Turner and Comfort in reading ENEBLEPSEN in the gap, but this is not really an argument at all. We simply do not know what verb may have stood after PETROS and in a short manuscript which such a free text (several singular readings) we cannot simply assume something is there in order to prove a connection with another text.
  2. Furthermore, there is no actual evidence that Marcion’s text read ‘Peter looked at him’. Indeed there is no ancient evidence that Marcion’s text contained anything of Luke 22.49-62. As Clavez notes, Epiphanius specifically notes that Marcion’s Gospel lacked the section about Peter and the ear of the high priest’s servant (p. 432 note 94). But she fails to note that there is no positive evidence for the presence of the following section (52-62) in Marcion’s version at all. The second page of P69 contains material that is not attested in Marcion (on p. 431 note 92 such a position is taken by Clivaz as evidence of Marcionite omission of material; she doesn’t register the inconsistency in her argument).
  3. Clivaz attempts to show that the reconstructed text of P69 at Luke 22.61 can be read in a Marcionite manner as undermining Peter’s status because ‘Jesus cannot bear to look at him’ (p. 432). This may well be an imaginative reading, but doesn’t correspond to the normal Marcionite practice which she uses to illustrate the situation here: the omission of material that doesn’t fit with the Marcionite profile (see p431 note 92, which seems to assume re Luke 22.32 and 23.49 that since ‘neither passage is attested in Marcion’ that is evidence of Marcion deleting this material).

Sixthly, a good deal of Clivaz’s case rests on the argument that a scribe, troubled by apologetic difficulties, may have deleted 22.42 by strategy of "negating the objections" by omission; and that this strategy ‘would be practical only in a type of Christianity that preserved a single gospel, as did Marcion.' (p. 429). I don’t see this argument as having much force. Firstly, there is no other evidence for the omission of 22.42 in the textual tradition of Luke. Secondly, there are practically no scribal practices that work consistently over the whole four gospel canon. Thirdly, the evidence (especially from Oxyrhynchus) shows decisively that the gospels were copied predominantly individually right through until the fourth century. Thus the particular connection to a Marcionite redaction doesn't follow. Perhaps we should also note that we have no other evidence of Marcionite material from Oxyrhynchus.

In conclusion I think the suggestion that P69 is a manuscript of Marcion’s Gospel is a very clever idea which is however not proven and not the most plausible context for making sense of this fascinating manuscript.


  1. what's the difference between an omission and an absence?

  2. Although Dr Head presents many helpful comments in regard to the alleged "Marcionite" nature of p69, at one point I think he overreaches:

    Head: "... the evidence (especially from Oxyrhynchus) shows decisively that the gospels were copied predominantly individually right through until the fourth century."

    Since the early papyri are mostly mere fragments, one normally would expect most of these MSS to display only a portion of a single gospel, regardless of original form and content.

    Even p39 (by its pagination) at best suggests that either (1) the MS contained only Jn; or (2) Jn may have been the initial gospel in some larger corpus.

    Certainly, when the more extensive pre-4th century papyri are examined, the evidence is clear: larger collections did exist, and (for all we know) may have been the norm rather than individual gospels circulating separately in that era.

    Consider the following:

    P6 – (g-k) Jn Jas (Saec.IV)
    (Did this MS contain the general epistles and Jn, the general epistles and possibly the four gospels or even Acts?);

    P45 – Mt Mk L Jn Ac (Saec. III);

    P53 – Mt Ac (Saec. III)
    (Did this MS contain the four gospels and Acts?);

    P4+p64/p67 – Mt Lk (Saec. II/III)
    (allowing for the sake of argument that Comfort/Barrett correctly regard p4 as part of the same MS as p64/67);

    P75 – L Jn (Saec. III)


    When one turns to the post-4th century era, where there should be no question that larger gospel collections would occur in papyrus MSS, the proportion of MSS demonstrating any such combination remains almost identical to that seen among the pre-4th century papyri:

    P2 – (g-k) Lk Jn ((Saec. VI)
    P44- Mt Jn (Saec. VI/VII)
    P84 – Mk Jn (Saec. VI)

    I would suggest that the data is either inconclusive or (in my opinion) more likely suggests that multiple NT books already were being combined in pre-4th century papyrus MSS.

  3. Also, since we have codices of the entire Bible, OT and NT already in the 4th c. Isn't it likely that 4-Gospel codices had already existed for awhile? The titles Kata Markon, etc., would seem to have arisen as means of distinguishing the 4 Gospels from one another in a single volume, using the attributed authorship of that means of distinction. This is very similar to the titles of the Pauline letter collection being according to recipient, since all are from the same author. And since the basic titles are so entrenched and ubiquitous in our extant mss, even in the earliest ones that are large enough to contain the beginnings and/or ends of the ms, it would stand to reason that 4-Gospel codices would have to have arisen early enough to leave such a mark on the recieved book titles. Irenaeus doesn't mention a 4-Gospel codex, but he does make a huge deal about there being exactly 4 Gospels. Likewise Tatian's Diatessaron was born out of a belief that these 4 Gospels naturally belong together. The uniting of the 4 Gospels as part and parcel of a complete record may even be witnessed in as old a work as the Shepherd of Hermas (see C. Taylor's 1917 monograph). I would tend to think, in light of the accumulation of this circumstantial evidence, that 4-Gospel codices existed by the early 2nd century.

  4. I had said: 'The primary characteristics of Marcion’s text are determined mostly be omissions and absences (minuses).'
    Anonymous asked: 'what's the difference between an omission and an absence?'
    Short answer: nothing really.
    Long answer: calling them 'omissions' tends to presuppose the patristic view that Marcion omitted material from a longer text. Calling them 'absences' is perhaps a somewhat more neutral way of describing them. Using 'minuses' borrows a terminology developed within LXX studies to describe the relation between the LXX and MT without presuming one omitted form the other.

  5. Maurice Robinson and Eric Rowe took exception to my comment that "... the evidence (especially from Oxyrhynchus) shows decisively that the gospels were copied predominantly individually right through until the fourth century."

    I think this is a little off topic. My view on this does seem strangely idiosyncratic at the moment (although I think it will become the general consensus in the future). I promise to defend my view fully at a future date (as I don't think it is that critical in my discussion of whether P69 is Marcionite).

  6. Am I allowed to fob respected blog commenters off like that, or do I have to answer the challenge?

  7. PH wrote

    "Am I allowed to fob respected blog commenters off like that, or do I have to answer the challenge?"

    fob ... off ?? I am going to try that on someone in seahurst park tomorrow and see what happens.

    If you will retract the prediction that it will become the consensus ... then we will let it go since it isn't central to your main point. Perhaps it is a good idea to avoid tangental controversial material in presenting an argument. Gives people something to nit pick about and TC folks are famous for nit picking.


  8. There is no point retracting the prediction. You will see it come true in your own lifetime.

    I agree about not introducing controversial points into an otherwise excellent (?) argument. The only problem is that I don't see it as controversial. Oh dear.

  9. OED (slightly ellided in my favour):
    "fob off. a. To attempt to satisfy with an excuse or pretence; ... to put off (a person) with (something of inferior quality or something less than he has been led to expect)."

  10. Head: "I think this is a little off topic....I don't think it is that critical in my discussion of whether P69 is Marcionite)."

    Had Dr Head of course not brought up that particular point in his original posting, the presumably "off-topic" comments would not have ensued. Ergo, the comments in fact are properly on-topic, even if related only to a separate side issue.

    As to reasons why this point does merit discussion:

    1) If pre-4th century gopspels on papyrus tended to travel in collections as opposed to independent circulation, then Marcion's use of Luke's gospel alone would represent a major departure from the "norm", and would ensure a strong "orthodox" reaction.

    2) On the other hand, if the gospels tended to circulate as individual entities during the first three centuries, then Marcion's alterations and actions might not have caused as large a stir as they did.

    Until the forthcoming definitive "consensus proof" regarding the latter alternative appears and convinces me, I will continue to presume that which I have previously stated.

  11. Although Marcion does seem to have paraphrase occasionally as well as omitted material, there is no suggestion of a connection between Marcion’s text and the D-type text elsewhere (I stand to be corrected on this point).

    I cannot find where Clivaz suggested the a connection between Marcion's text and the D-type text in her article, but it asserted by other critics. For example, Friedrich Blass, The Philology of the Gospels (Macmillan, 1898): 145-146 wrote: "Now it cannot be doubted that Marcion's text of Luke's Gospel did not exhibit the form commonly known, but that attested by other Western evidence, such as D and Latin versions, and in this way a different Western text of the Gospel is probed to have existed as early as in the first half of the second century."

    The quadruple negation ("cannot", "doubt", "did not", and "but") is confusing, but based on the context Blass is asserting that Marcion's text is like D. (NB: Blass's "the form commonly known" is the Alexandrian text.)

    Granted, Blass is over a 100 years old...

  12. Thanks Stephen for picking up an important point:

    SC: "I cannot find where Clivaz suggested the a connection between Marcion's text and the D-type text in her article, but it asserted by other critics."

    Clivaz didn't suggest this connection, but it is implicit in her proposal. P69 has some affinities with the D-text. I was suggesting that the lack of affinity between Marcion's text and a D-text could be held against her proposal.

    My statement, that there is no connection between Marcion and D, was based on the silence in my head on the subject (dangerous, yes I know - hence 'I stand to be corrected' - which indeed I am being!). I only checked Parker on Bezae who doesn't note any connection with Marcion. Scrivener I now find the same, no mention of Marcion.

    BUT. From the other side, Harnack suggests that there is a special affinity: 'Speciell aber dem Text im Cod. D steht M[arcion].s Text na/her als jedem anderen Text.' and in a note to that: 'Die besondere Affinita/t D und Marcion ist im Ev. noch gro/sser als im Apostol.' (Marcion, 223* - sorry, can't figure out umlauts in the comments) And, what is more, he provides evidence for this which looks impressive.

    Marcion notes that the Old Latin archetype behind a b c e is even closer to Marcion than D.

    So presumably Blass is working on the basis of the same evidence in his point about 'D and Latin versions' (as Stephen noted in the previous comment).

    OK, this is good. Learning things etc. And this point as relates to Clivaz should probably be withdrawn. But it raises other big questions as to the nature of the 'affinity' and why I didn't already know this. Just ignorance? [probably] Or a datum that has been somewhat lost sight of in recent years? [still amounts to ignorance I realise, but sounds more excusable] Maybe there are books and articles out there on this.

  13. Where is Ulrich when we need him?

  14. Eric, can you please give more details of C. Taylor's 1917 monograph? Thanks.

    Also, it would be interesting to have a list of the aspects of Hermas (specific texts) that made Turner surmise that the book attested a fourfold Gospel.

  15. Marcion [Harnack?] notes that the Old Latin archetype behind a b c e is even closer to Marcion than D.

    That's pretty interesting. According to a collation of P69 I did, P69 tends to support the OL (incl. ff2) over D where the OL and D differ.

    (NB: Blass is reliant upon Zahn's reconstruction of Marcion's text in his Geschichte des N.T. Kanons 2.2.1, p. 411ff.)

  16. "Eric, can you please give more details of C. Taylor's 1917 monograph? Thanks."

    Oops. I was way off on the date. It was actually published in 1892. I'm not sure what I was thinking of. And the title is Witness of Hermas to the Four Gospels. Unfortunately our library copy is out right now. But I believe that J. A. Robinson's commentary on Hermas advocates the same view. There is a passage in Hermas that contains some striking verbal parallels to material from the canonical Gospels and that plays up the number 4 in way that seems to betray a numerological significance to it that Hermas doesn't spell out. If I remember right, some of his language about the number 4 is similar to Irenaeus' defense of the need for 4 and only 4 gospels. So the case doesn't really come down to any firm proof, but its still a suggestion that I think comports very well with alot of the mysterious nature of the symbolism in Hermas. I'll try to hunt down either Taylor or Robinson so I don't have to go by my memory (which is clearly pretty shaky on this as evidenced by my 25 year error in the date).

    Then again, I have a hunch that Michael Holmes could write a whole blog entry on this very topic that would be far more informed than what I could put in a comment. What say you Dr. Holmes?

  17. From an electronic scan, there is not much in Hermas that could be taken as relating the number four to the canonical gospels. The only passages that might suggest such are symbolic, paralleling as suggested the comments of Irenaeus. These are the following (LCL translation):

    Sim 9.4.3 ...there became four tiers in the foundations of the tower...

    Vis 1.4.1 ...there came four young men, and took up the chair and went away towards the East...

    Vis 3.10.1 ...and four others took up the couch and bore it away also to the Tower...

    Vis 3.13.3 ...a couch has four feet and stands securely, for even the world is controlled by four elements...

    That seems to be all.

  18. The following is copied from J. Armitage Robinson's Barnabas, Hermas, and the Didache (1920) regarding Vision 3.13 (except for the quote from Hermas which I copied from Bibleworks rather than Robinson):

    "The explanation of the final form in which the Church has appeared contains certain interesting allusions which were first pointed out by the late Dr. C. Taylor. The words are as follows:
    'For, just as when some good news (aggelia agaqh tis) comes suddenly to one who is sad, immediately he forgets his former sorrows, and looks for nothing else than the good news which he has heard, and for the future is made strong for good, and his spirit is renewed on account of the joy which he has received; so ye also have received the renewal of your spirits by seeing these good things. 3 As to your seeing her sitting on a seat, that means that her position is one of strength, for a seat has four feet and stands firmly. For the world also is kept together by means of four elements.'
    "In the words 'immediately he forgetteth his former sorrows,' and 'his spirit is renewed for the joy that he hath received; so also ye... by seeing these good things,' we have a clear allusion to St. John xvi. 21: 'A woman when she is in travail hath sorrow, because her hour is come: but as soon as she is delivered of the child, she remembereth no more the anguish, for joy that a man is born into the world. And ye now therefore have sorrow,' etc.
    "But there is more than this: the couch on which the Church sits, when she is thus revived by the good tidings, is firm because it has four feet; and the four feet corresponde to the four elements of the world. Dr. Taylor reminds us of the famous words of Irenaeus (III. xi. 11) about the Four Gospels, and the strange parallels which he brings to show that there must needs be four and no more: moreover he points out that the allusive method of Hermas makes it reasonably certain that 'aggelia agaqh' is his way of avoiding the obvious word 'euaggeliou.'"

  19. Since Hermas is a highly symbolic book that often leaves it up to the reader to decode some of the symbols despite superficially giving explanations for the visions, I find Robinson's interpretation of the significance given to the number 4 here reasonable.

    First, the text deliberately explains why the fourfold foundation of the Church must be precisely fourfold (whatever that foundation may be). This means that the number is not an incidental detail in the vision, but something that represents some real fourfold thing that Hermas views as foundational to the Church.

    Second, the explanation given for the fourfold foundation of the Church does resemble Irenaus' argument for there being 4 Gospels in Adversus Haeresus 3.11.8:
    "The Gospels could not possibly be either more or less in number than they are. Since there are four zones of the world in which we live, and four principal winds, while the Church is spread over all the earth, and the pillar and foundation of the Church is the gospel, and the Spirit of life, it fittingly has four pillars, everywhere breathing out incorruption and revivifying men."
    Hermas' vision also includes appeal to the 4 winds, the idea of being a foundation to the Church, and revivifying qualities.

    Third, as Robinson points out, a verbal parallel to one of the canonical gospels is tucked within this section.

    Fourth, Hermas connects the mysterious fourfold foundation of the Church to good news (aggelia agaqh). Seeing this with Robinson and Taylor as a circumlocution of euaggelion is a way to make sense out of a section of Hermas that would otherwise opaque. And while it might be objected that "opaque" is Hermas' middle name, I tend to think that, although the Shepherd is pretty hard to understand, he always has a purpose for his symbols and doesn't throw them together haphazardly.

    Are we far enough off topic yet, Pete?

  20. Eric, you mean Armitage Robinson, not Maurice Robinson, right?

  21. Eric: "Are we far enough off topic yet, Pete?"

    A long way from P69 and Clivaz's article, but the subject of when the fourfold gospel arose is surely a central question to TC.

  22. "Eric, you mean Armitage Robinson, not Maurice Robinson, right?"

    Yes. Armitage.

  23. PJW said: 'the subject of when the fourfold gospel arose is surely a central question to TC'.

    Yes, but perhaps it would help if we were clear what we meant by 'the fourfold gospel'.

  24. Dr. head,

    I think as a witness to Marcion is a reasonable conclusion. As the Marcionite text is not certain. If it is not, then it means the prayer is a very late Orthodox corruption.

    The reasons to remove the prayer are non-existent for a Catholic scribe. You simply cannot find a theology that would logically do so, except Marcionite.

    Of course this gets back to your basic problem, which is you think Christian texts lived for 100 years undetected, and then when Marcion and others hit the scene they deleted material.

    Of course Marcion was an incredible expert, erasing every occurrence of TE (the theology of that word was obviously unacceptable) as well as other theologically loaded words like PARACRHMA. Frankly that is where you case breaks down, on the vocabulary. There are dozens of words missing from Marcion completely, (mercy, reckon, etc) which would be an impossible task to remove from 11 books.

    You seriously need to re-examine your assumption that Marcion erased text. Tertullian was very wrong.

    1. It could also be that Marcions "discovery" of Galatians and three other of Paul's texts was a creation of those writings.

  25. Despite the Irenaeus’ assertion, Eric Rowe, the Church may revivify men with only three pillars… : see Galatians 2, 9 !

    Reading ΠΕΤΡΟϹ in 22:61 : I am not convinced at all ! yes, we have the article Ο (?) – it seems –, but look at the verso – recto is useless – : all Ε following an “horizontal” letter (Π, Ϲ, Τ) is written high, raised, with its middle stroke at the same level than the preceding letter ; compare ΠΕ, ϹΕ and ΤΕ with the other pairs ×Ε. Compare ΦΩΝΗϹΕ, ΦΕΙϹΟΠ× and ΤΩΤΟΤΕ… So, if ΦΕΙϹΟΠ××× is not [ϹΤΡΑ]ΦΕΙϹ Ο ΠΕ̣Τ̣Ρ̣[ΟϹ] ?… « [×××στρα]φεὶς ὀπ[ίσω] », maybe ? What else ? And maybe ho Petros in 22:61a means its omission in 61b :

    [22] ᵃφεις ο πε̣τ̣ρ̣[ος ενεβλεψεν αυ]
    [15] τω ᵇτοτε [υπεμνησθη ο πετρος]
    [22] τ̣ου ρημ[ατος του κ̅υ̅ ως ειπεν]

    [Peter M. Head : « Fifthly, (…) 2. Furthermore, (…) As Clavez »… : “Clivaz” ?]