Our two oldest witnesses to the Greek text of Paul support this reading: the Chester Beatty Papyrus of Paul (P46) from around AD 200; and Vaticanus from the fourth century. This early Alexandrian textual tradition is supported in a strong group of 'Western' witnesses (D G it etc.).
P46, the earliest evidence for the reading QEOU KAI XRISTOU in Gal 2.20.Some account of the slender argumentative basis behind the general consensus of contemporary scholarship can be seen in the arguments summarised in Metzger's Textual Commentary:
- God and Christ ‘can scarcely be regarded as original since Paul nowhere else expressly speaks of God as the object of a Christian’s faith’ (p. 524).
- ‘the son of God’ best explains the origin of the other readings:
- ‘It is probable that in copying, the eye of the scribe passed immediately from the first to the second TOU, so that only TOU QEOU was written (as in ms. 330); since what followed was now incongruous, copyists either added TOU UIOU or inserted KAI XRISTOU.’ (p. 524)
- Romans 4!!
- Metzger's argument involves an imaginary process, a crucial part of which is found only in a single twelfth-century manuscript. It doesn’t explain what may have drawn a scribe to add ‘and Christ’, although it does seem to acknowledge that this was a very early variant.
- It is really a desperate measure because as is universally admitted the ‘Son of God’ reading is a smooth reading, making perfect sense and generally consistent with Pauline thought. If one asks what sort of reading scribes would have preferred clearly the answer is precisely this sort of reading.
- Here the more difficult reading seems pretty clear and it has very good external attestation.
So what Paul wrote was obviously (!): 'The life I now live in the flesh, I live by faith in God and Christ who loved me and gave himself over for me.'
Any thoughts?
Andrew Wilson believes that this reading arose in the Western tradition, and that p46 and B show a Western influence here.
ReplyDeletehttp://nttext.com
/forum/index.php?topic=58.0
Peter,
ReplyDeleteConcerning the syntax, the reading you have chosen is very weighty on the right hand side of KAI and I was wondering if an example of this could be found in "Paul". 2Tim 4:1 was the best I could find.
GAL. 2:20 ... EN PISTEI ZW THI TOU QEOU KAI CRISTOU TOU AGAPHSANTOS ME KAI PARADONTOS HEAUTON HUPER EMOU.
2TIM. 4:1 DIAMARTUROMAI ENWPION TOU QEOU KAI CRISTOU IHSOU TOU MELLONTOS KRINEIN ZWNTAS KAI NEKROUS, KAI THN EPIFANEIAN AUTOU KAI THN BASILEIAN AUTOU:
I agree with you that the "story" which was dreamed up, long before Metzger wrote about it, to explain this complex process sounds dubious. It is an example of a compound hypothetical situation where there are multiple "ifs" and the second and third ifs are dependent on the previous ones. Each new if in the chain makes the story more improbable.
clay
Dear Peter,
ReplyDeletea good thought. Tregelles
proposed the same reading.
On the other side, "live to God & Christ" may well be a combination of V19&20, where you have already V19 "living to God" and V20 "Christ lives in me"; so the B-P46 reading may be a harmonization to the immediate context, which is in face of the western attestation not easily dismissable.
Martin
Peter is wrong to call the readings of P46 and B Alexandrian here. P46 is much more Western in Galatians than Alexandrian, and B also has inherited a very strong Western element in Galatians (see nttext.com/forum for a table of manuscript relationships for Galatians (they are listed under Galatians Preparations)).
ReplyDeleteThe fact that the only other witnesses for Peter's preferred reading are D, F, G; It(b), d, g and Victorinus-Rome shows that this reading is a purely Western reading with no breadth of attestation.
IMO, the reading is also clearly clumsy and unlikely on the grounds of what Paul was driving at in the verse. You may find my comments on the verse under Galatians Ch. 2 on the forum.
Tregelles had not even the advantage of P46!
ReplyDeleteInterestingly, Tregelles cites the reading of Victorinus as 'in fide vivo filii Dei et Christi'. This seems to be in conflict with the apparatus in NA27. Some will say it is a conflation of the two readings, others will say it explains best their origin.
In NA27, a critical mark remains above Ga 2:20 AGAPHSANTOS, yet with nothing in the apparatus.
ReplyDeleteI presume that NA27 intended to eliminate the questionable Marcionite reading AGORASANTOS cited in NA26 (Marcion per the translation of Rufinus; cf. the discussion in U. Schmid, Marcion und sein Apostolos, 232-233).
Muenster needs to remove that angle bracket in subsequent printings.
As for the primary variant under discussion, PMH writes:
"Metzger's argument involves an imaginary process, a crucial part of which is found only in a single twelfth-century manuscript."
Metzger certainly does not claim MS 330 as the archetype, but cites such as an example of a possible hom. error that could easily have occurred during the (undocumented) pre-transmissional history within Western sources. Contra CSB and PMH, however, such speculation is not in itself illegitimate, since textual critics of every stripe have made similar appeals elsewhere. The primary issue is whether the speculation is reasonably plausible, and in this case I think it is.
On the other hand, I am puzzled by PMH's further statement:
"It doesn’t explain what may have drawn a scribe to add ‘and Christ’".
?? -- It seems to me that Metzger explained that quite well. Had TOU UIOU had been omitted (in a pre-Western archetype) by hom. of TOU...TOU, the text would read "the God who loved me and gave himself for me". This in itself would prompt some sort of amplified correction, of which the addition KAI CRISTOU might be the most logical ("...God, and Christ who loved me...").
Metzger's scenario here seems quite plausible to me.
MAR, the critical mark is not in my printing of NA27 (post 2001).
ReplyDeleteThis stream of NA27 corrections is starting to remind me of the "WHICH King James Version?" debate.
ReplyDeleteNA27 already indicates where it diverges from NA25; how about an endpage in each printing showing how it keeps diverging from itself?
Something to keep in mind for those preparing NA28.
Andrew said: 'Peter is wrong to call the readings of P46 and B Alexandrian here.'
ReplyDeleteOne of the interesting things about this variant is precisely the agreement between 'Western' witnesses and (normally) good 'Alexandrian' witnesses. This situation is not peculiar to this variant of course, but poses an interesting general methodological question as to what to make of these type of readings. Of course one can think of them (with Andrew apparently) as Western contamination of the Alexandrian witnesses at precisely these points. (Did Zuntz take this view? my memory fails me and I wasn't able to access the blog from my office all day.) But that is not the only possibility, as one can think of these as having a point of origin shared by both Western and early Alexandrian 'types'. Or, as being the original reading.
Peter Head wrote,
ReplyDelete"Of course one can think of them ... as Western contamination of the Alexandrian witnesses at precisely these points. (Did Zuntz take this view?)"
No, in fact this view is essentially that of the UBS Editorial Committee, which repeatedly speaks of "western" contamination in Alexandrian witnesses.
Zuntz sees matters quite differently: in his opinion, the “outstanding feature” of the group of witnesses he termed ‘proto-Alexandrian,’ and in P46 most of all, is the presence of
"‘Western’ readings, or rather, those readings which have disappeared from the later ‘Alexandrian’ manuscripts (and often also from other Eastern witnesses) but recur in the West. The presence of these readings does not make the group ‘Western’ in any legitimate sense of the term; the ‘Alexandrian’ character of the ‘proto-Alexandrian’ witnesses is established by unequivocal facts. This element, common to the earliest Eastern and to the Western traditions, is a survival from a pre-‘Alexandrian’ and pre-Western basis, the traces of which, most marked in P46, gradually disappear from the later ‘Alexandrian’ tradition but often reappear in later Eastern witnesses, as well as in the West" (Text of the Epistles, 156-7).
In other words, "Western readings in non-Western witnesses are, generally, ancient survivals. The are not, in the relevant witnesses [e.g., P46], secondary intrusions into a previously pure form" (Ibid., 142).
On this point I'm persuaded that he is essentially right, and in two articles due out sometime between May and October attempt to develop the implications of his position.
I agree entirely with Mike about Zuntz in the broad sweep of things.
ReplyDeleteHowever, my point was that in Galatians P46 is not Alexandrian in any discernable sense (except the fact that it was found in Egypt).
For example, in the 25 most significant variants in Galatians (those listed in the app. of UBS4 in which 2 or more mss agree against the rest), P46 agrees with 01 only 8 times, A 5 times, and C only 4 times. In contrast, P46 agrees with F (and G) 10 times and D 8 times.
P46 agrees most with B (13 times), and even B has as it closest relatives 01 (13 agreements), followed by A (12), D (12) and F, G (11). The relationships of all the major mss are not exactly easy to describe in the normal 'text-type' terms.
Maybe the sample size for Galatians could be a bit larger. But, these figures demonstrate that Mike might be over-stating matters slightly when he says that 'the ‘Alexandrian’ character of the ‘proto-Alexandrian’ witnesses is established by unequivocal facts'.
From (my admittedly poor) memory, Zuntz does not give any tables of manuscript relationships for the Pauline corpus, let alone for individual books. Maybe he believed that the books only ever circulated as a corpus and so generalized the results?
As far as Gal 2:20 is concerned, we appear to have three choices:
(a) the reading is a Western intrusion into Alexandrian witnesses (contra Zuntz)
(b) the reading is an ancient survival in Alexandrian witnesses (with Zuntz)
(c) the reading is an early Western reading in a non-Alexandrian witness (P46), from which the later witness (B) is a descendant.
If P46 is not Alexandrian in any textual sense, then only option (c) remains viable.
One clarification: the statement that the " ‘Alexandrian’ character of the ‘proto-Alexandrian’ witnesses is established by unequivocal facts' " is Zuntz's, not mine.
ReplyDeleteThanks Mike for the clarification on Zuntz (I call it a 'clarification' because otherwise I'd have to use a form of words which might make it seem as if I had been 100% wrong). I'm going to have to read Zuntz again! (Anybody know anything about the proceedings of the Zuntz 50th anniversary conference by the way?). Obviously I am with Zuntz, not against him.
ReplyDeleteThanks Clay for the 2 Tim 4.1 parallel. That is quite helpful as illustrating a somewhat parallel construction and a very parallel piety.
ReplyDeleteThus another point in favour of this reading in Gal 2.20 is that it supports the authenticity of 2 Timothy.
As well as Tregelles, Bover (1953) has this reading.
ReplyDeleteMartin said: 'On the other side, "live to God & Christ" may well be a combination of V19&20, where you have already V19 "living to God" and V20 "Christ lives in me"; so the B-P46 reading may be a harmonization to the immediate context, which is in face of the western attestation not easily dismissable.'
ReplyDeleteThis is an interesting thought. My initial response would be that although this is better than Metzger's explanation for the origin of this reading, it is not compelling. It seems to me that most harmonisations to context are more literal and less conceptual than what you are proposing.
I think I could use this argument, especially from v19 in support of the reading as Pauline: the bidirectional life of which Paul speaks is clarified as a life of faith with two foci: God and Christ.
Andrew said: 'my point was that in Galatians P46 is not Alexandrian in any discernable sense (except the fact that it was found in Egypt).'
ReplyDeleteI take it from your later comments that you are probably envisaging this as applying particularly (or only) to Galatians within P46; i.e. that different books might reflect different text-types and thus provide remnants of earlier, separately transmitted conditions.
I would love to see this point worked up and proven (something like this has been argued w.r.t. Rom 16 in P46 some time ago). Perhaps Colossians would be another good case for exploration as it has an unusual concentration of variants.
But, Andrew further said: 'For example, in the 25 most significant variants in Galatians (those listed in the app. of UBS4 in which 2 or more mss agree against the rest), P46 agrees with 01 only 8 times, A 5 times, and C only 4 times. In contrast, P46 agrees with F (and G) 10 times and D 8 times.'
I think this is not for me a significant argument. UBS4 is simply too selective to use in the attempt to categorise the text-type/group/familie of a manuscript. All it would prove is that on the occasions when P46 does agree with DFG then it has some plausibility and some significance for translators.