Tuesday, September 19, 2006

Thoughts on Singular Readings

11
As relatively new to this blog I would be interested to know how others here view the appropriateness of singular readings among GNT manuscripts for an eclectic NT text? (Forgive me if this is already discussed somewhere on the blog. I am not referring to using singular readings for understanding a manuscript's character.)

Does anone have a favorite example as a contender for an acceptable text?

Do any/most/all accept as a principle that a singular reading should not be followed in an NT eclectic text? This would imply, qal vaHomer, that emendations should not be accepted in an eclectic NT text.
(PS: I accept that a manuscript could include good, authentic information, e.g., Bezae, without implying that such was part of the author's original.)

So, as a curiousity, what do people on the blog think/do? Or how far are they willing to take a principle?

11 comments

  1. For example, NA27 at Lk19:38 reads EULOGHMENOS O ERXOMENOS O BASILEUS EN ONOMATI KURIOU
    As far as I can tell, this is only attested in one Greek manuscript B-Vaticanus G03. A singular reading.

    Obviously, a committee has accepted it. And it would be an exception to a tight, externally-principled eclecticism.
    Who is happy with NA27 here? Who isn't?

    Anyone have a list of others?

    ReplyDelete
  2. I'm not happy with NA27 here. I prefer 2 as a minimum number of witnesses, except in the Apocalypse. I think that one should be 'conservative' about what one puts in the main text of a Bible: conservative towards speculative translations, conservative about retroversions, conservative about readings that might only turn out to be the result of the scribe of the ms itself, and conservative about readings that are only found in late mss. This conservatism with regard to what you print does not apply in the same way to theories that scholars explore. But we shouldn't use a printed text of the Bible to explore our own pet theories.

    ReplyDelete
  3. Even if you disliked following a single Greek witness I don't think that Luke 19.38 should trouble you. Of course it depends on how you delimit the variation units. The whole quoted passage may be singular to B, but if you break it down to the fundamental question as to whether we read O BASILEUS or BASILEUS then we find that 01, 03 and 05 all read O BASILEUS. On that basis I'm perfectly happy with NA27 text here.

    ReplyDelete
  4. PJW wrote: 'I prefer 2 as a minimum number of witnesses, except in the Apocalypse.' So would I be right to gather from this that there is actually no text critical principle involved other than some innate conservatism that somehow doesn't apply in Revelation.

    ReplyDelete
  5. I personally don't see that I should disallow singular readings as 'contenders' as a general principle. In fact, however, depending on a sole witness would make me pretty cautious, especially if the singularity could be explained as the product of a pattern in the character of the particular manuscript (as e.g. Heb 3.10b contrast Attridge's preference for the singular reading of P13).
    I also wouldn't limit myself to Greek manuscript witnesses in defining singularity in this sense (e.g. cf. Mark 1.41 with only D but some Latin support).
    I can't see myself voting to print a conjectural reading (although I accept that we do often use conjecture in the construction of 'the reading which explains the others'), although there are places where I would certainly entertain conjectures for problems that no witness seems to solve (e.g. Eph 1.1). And would like to see them in the apparatus.

    ReplyDelete
  6. PMHead: "In fact, however, depending on a sole witness would make me pretty cautious,"

    Good. Me too!

    PMH: "especially if the singularity could be explained as the product of a pattern in the character of the particular manuscript"

    Especially!

    PMH: " ... also wouldn't limit myself to Greek manuscript witnesses in defining singularity in this sense (e.g. cf. Mark 1.41 with only D but some Latin support)."

    wow. Following D in a singular reading is the ultimate in gutsiness (!), probably the most prolific source in attractive singular readings and rejectable singular readings.
    As much as I like some of D's readings, and dislike others of D's "special readings", I lean towards ennunciating a principle that D should not be followed alone. Does Latin support provide sufficient support? Hmmm.
    I'm giving a paper in Nov where among other things I will reject D's reading on Mk 15:34 even though I think it historical. Is it cold feet or do I just not trust D's character? Truly a great manuscript to read, but to establish the text on it? (To be fair, it is ultimately my reading of Mark's literary pattern and purpose that give me the confidence that D's reading was not Mark's writing.)

    Another classic spot: Acts 15 and the apostolic halaxa. "D" fits understandings of 1st century Judaism and provides a sensible historical reading. But if Luke wrote "D", where did eveything else come from? I'd like to follow 'D', but out of principle go with 'Alexandrian'.

    ReplyDelete
  7. My favorite singular reading is D's tidy omission of the extra Cainan in Luke 3:36.

    As much as I wish it were original, the testimony of D in this passage carries less weight for establishing the original text of Luke than the inscriptions discovered by Ron Wyatt do for interpreting Genesis 8 and Exodus 14.

    Unlike many textual critics of today, however, I read D's omission of Cainan as evidence for the text of the LXX as it stood before Origen got his hands on it.

    ReplyDelete
  8. Re the number of witnesses for a reading: let's not forget that the number of extant MSS supporting a reading today may not bear any relation to the number of MSS in antiquity supporting that reading. E.g., Eusebius reports that "nearly all the copies of Mark" end at 16:8, a report echoed by Jerome, who observes that the "longer ending" is "found in only a few copies of the Gospel--almost all the Greek copies being without this final passage." Today, of course, only 3 Grk MSS end at 16:8. Similarly, regarding Matt 5:22, Jerome notes that "most of the ancient copies" do not contain 'without cause' (eike), which is found in the great majority of MSS today. In the other direction, in Rom. 3:9, what Arethas of Caesarea reports as the reading of the oldest and most accurate MSS (katechomen perisson) is a variant not found in any extant MS.

    So (agreeing with Peter Head) in principle, singularity alone cannot be a bar to consideration of a variant. At the same time, it would require, for me at least, a heavy preponderance of all other sorts of evidence in its favor for me to give it serious consideration.

    ReplyDelete
  9. The singular reading for which NA27 is most infamous in my eyes is A's omission of 'hmas' from Rev. 5:9.

    I'm not certain what the sum total of evidence against that omission is, but Hoskier compiled quite a list, noting that no other mss or quotation in any language other than Ethiopic, supported this omission. He even notes under #7 below those ms of Revelation that are defective in loc, to which of course must be added p24 and p47.

    Apparently, all of the Greek mss support Variant #3 (Erasmus 4-5) except as noted below (the 4th reading is that of Erasmus 1-3):

    1. TW QEW: 02 eth
    2. TW QEW HMWN: 180 205 1854vid 2053 2079 2082* 2256vid
    3. TW QEW HMAS: 01 046 209 1006 1352 1611 1841 1859 2020 2042 2053 2065c 2081 2138 2329 2351 2432 Maj cop Andreas(a,p) Arethas
    4. HMAS: 296 2028 2029 2033 2044 2054 2065* 2068 2069 2070 2083 2186 2305 2814 it(z*) vg(ms) arm1 Iranaeus(lat)vid Cyprian Fulgentius
    5. HMAS TW QEW: 94 104 172 250 254 336 424 459 469 582 616 620 628 680 792 922 1746 1828 1862 1888 1918 2016 2018 2019 2032 2038 2050 2073 2084 2254 2344 2595
    it(ar,gig) vg syr(ph,h) cop arm arab Hippolytus; Cyprian Maternus Augustine Varimadum Fulgensius Primasius Beatus
    6. HMAS TW QEW HMWN: 205 cop sa arm3
    7. Hiant: C P 051 88 1384 1704 1774 2022 2030 2062 2078 2091

    ReplyDelete
  10. PMH, I would want to distinguish between 'allowing singular readings as contenders', i.e. being willing to envisage the possibility that they might be original, and the editorial decision to print them. I do not have a text-critical argument that could demonstrate that all singular readings are secondary. However, when an editor of the GNT puts something in the main text they are making a statement about probability.

    Now whatever internal grounds there are for prefering the reading of B in Luke 19:38 they are, to say the least, slight. Preference for B's reading rests on marginal judgements of probability of transmission during stages to which we have little access. Can these outweigh the considerable possibility that B's reading does not go deep into its textual pre-history?

    Editors should be conservative, but good editorial procedures, with built in caution, should not be confused with arguments for truth.

    However, at a theological level I would also like to raise the issue of our role. I have earlier argued (earlier post, somewhere!) that if God has seen fit not to preserve the original text for us at some point, that is his business not ours. I wonder if we could extend the principle to say that it only becomes incumbent on us to consider a reading when the manuscript support for it becomes sufficient (a term I will leave here undefined).

    ReplyDelete
  11. Buck: Unlike many textual critics of today, however, I read D's omission of Cainan as evidence for the text of the LXX as it stood before Origen got his hands on it.

    I would hardly consider this argument from silence to be a better or more plausible solution than simple homoioteleuton, skipping from TOU to TOU.

    Also, the possibility that p75vid may have made the same error serves only to accentuate the issue.

    ReplyDelete