Monday, October 10, 2011

The Exemplar of P46

P46 is a manuscript copied from a Greek - Latin exemplar. This explains the many readings of P46 that are subsequently found in later Greek - Latin manuscripts and it explains the direct influence of Latin errors in the Greek.

[Don't quote me yet on this, please read on.]

In that doorstop of a study Scribal Habits, James Royse discusses the singular reading επευξ of P46 in Heb 5:6. This is almost certainly nonsense as it stands, having replaced the normal ιερευς ('priest'). Zuntz suggested that this reading is a Latin alphabet error, a replacement of Greek Rho ρ for a Latin 'p' which is then graphically represented as a Greek Pi, π. The Xi ξ is a 'simple' replacement of ς. Royse rejects this explanation ('there appears to be no other evidence that our scribe was in any way influenced by Latin') and treats the reading as an inexplicable error (though he treats the preceding ει as a separate variant).

First, I believe the whole variant is ειεπευξ for ιερευς. That is, the first syllable ει represents an itacistic reading of the initial iota of ιερευς.
Secondly, I may have found another instance of a Latin misreading in 2 Cor 10:12 where we have the puzzling νεκρουντες ('died') for μετρουντες.

The appearance of νεκρ- for μετρ- might be interference of the Latin stem MORT- as in MORTALES. Admittedly, a few other errors have to take place at the same time τρ - ρτ, but that is not an argument against this proposal. Τhe fact that we have a strange reading implies that a few errors have occurred—what is left for us is to reconstruct these errors.
Thirdly, I actually quite like Zuntz's suggestion, and would love to see it as the hypothesis for an extended study of P46. What if there is a large Latin element in this manuscript? Can other readings be explained by means of Latin influence? In addition, it would open up interesting questions on the 'how?' of the countless instances where P46 joins the Greek - Latin bilingual manuscripts virtually on its own.

So, this is the justification of my first paragraph. A theory that may be worthwhile to put to the test, either to reject or accept.


  1. But how likely is it that such a bilingual copy was in use in Egypt c. 200? All such mss from Egypt are much later (300-) , and were designed to help Greeks learn Latin to further their legal education.
    But cf. Chester Beatty Codex AC 1499 ed. Wouters (LDAB 3030, fifth AD), which has a Greek Latin Glossary of terms from the Pauline Epistles.
    See Cribiore, "Higher Education in Early Byzantine Egypt: Rhetoric, Latin, and the Law.” in Egypt in the Byzantine world, 300-700 ed. Roger S. Bagnall,

  2. Before I say anything else, your opening sentence says "many readings." But the rest of the post only has two. How many is "many"? And is there a list of them anywhere?

  3. On current evidence this is quite unlikely. Yet ...
    I guess, few would want to argue against the existence of at least one African Latin translation of Paul before the date of P46. Which means we have at least some people who have a Greek and a Latin version of Paul together (and how closely together - as within one particular MS - is a different question, though Origen produced his Hexapla around this time). Granted the later examples are for educational purposes, we have of course also the Greek - Latin inscriptions (Cyrene) from this time with a completely different purpose. [This only to demonstrate that bringing two languages together does not have to mean that we are talking 'schools'.]

    Eric, the two readings in the post are about Latin language/alphabet influence on the Greek text of P46, the 'many readings' refer to the P46 D F G variants you can find in the apparatus of your GNT.

  4. I don't think 2 Cor 10.12 was our scribe's finest hour, there seems to be multiple other things going on in the rendering of this verse in P46.

  5. Hoskier, as should be well known, had claimed that bilingual and even trilingual MSS may have been in existence during the 2nd century.

    Although his suggestion of trilingualarity might be seriously questioned, there is no good reason why 2nd century bilingual exemplars may not have existed; this, assuming the common view that the Old Latin originated during the early or mid-2nd century.

    Should this be the case, then no surprise if certain Old Latin readings (or misreadings!) should have affected the companion Greek side, or vice versa. The same would then apply to any purely Greek MS that may bave been copied from such a bilingual MS.

    So the possibility of Latin-based corruption at a couple of locations within p46 is not necessarily too wild a speculation. More such examples would be helpful, however, if one were attempting to seal the point.

  6. I would prefer to think at this point that a psychological interference (perhaps bilingual in nature) has contributed here without the necessity of postulating a Vorlage in Latin.

  7. I suppose, if you wanted to carry this further, you should look at Bilingualism and the Latin language by James Noel Adams.

  8. But the only way you can really be confident about this is to have the full profile of P46 first, in terms of the kinds of "errors" and corrections present in the extant 172 pages, and isolate those that agree with D and F, and validate whether this proposal is attested by the profile.

  9. Does the P46 version of Rom 16:15 betray a poor understanding of Greek conventions? Royse himself suggests that the scribe (or an earlier scribe) misunderstood the purpose of the letters alpha and beta written above the names "ΙΟΥΛΙΑΝ" and "ΝΗΡΕΑ", thus giving rise to "ΑΟΥΛΙΑΝ" and "ΒΗΡΕΑ". Royse's suggestion is that the Alpha and Beta had been inserted above the text as an indication that the name order should be reversed. See my post here.

  10. Several years ago, I noted that P46 appeared as a prime candidate for the source for the original Old Latin text. (see essay "On the Origin and Value of the Western Text-Type as Concerns the Pauline Epistles" available at Certainly the Old Latin began as an early translation from early Egyptian papyrus/papyri. P46 was found to have a considerable amount of material which was incorporated into an early Latin translation, that is it was used to create the first OL texts! (Either P46 or its earlier exemplars). Glad to see this present theory emerge!! However checking or comparing with codices D F or G (in the Pauline Corpus) is not recommended, as these bilinguals are highly contaminated. Better to refer to earlier more pure OL MSS, such as Beuron 64 and 79, et al. Good information is also seen in my "A Comparison of 10 Latin Manuscripts via Three Test Passages", also on the same website (s.v. "Versions") All of which opens a large door for future research. Exciting!
    Mr. Gary S. Dykes